Home Blog

ISO/IEC 42001 vs ISO/IEC 27090:2026 – Why AI Governance and AI Cybersecurity Are Not the Same Thing

Does the new AI cybersecurity standard replace ISO/IEC 42001, complement it, or expose an important distinction that organizations need to understand?

By Alexandros Engelen | Cybersecurity Strategist specializing in Maritime Cyber Risk


A new standard, and an important question

On 14 October 2026, the ISO – International Organization for Standardization and the IEC (International Electrotechnical Commission) published ISO/IEC 27090:2026, Cybersecurity – Artificial intelligence (AI) – Addressing security threats and compromises to AI systems.

The publication arrives as organizations across virtually every industry are expanding their use of artificial intelligence. AI is becoming embedded in customer services, cybersecurity operations, engineering analysis, business decision-making and critical infrastructure.

But the growing adoption of AI has also introduced a governance and security challenge.

How can an organization demonstrate that its use of AI is both responsibly managed and adequately protected against AI-specific cybersecurity threats?

ISO/IEC 42001:2023 already provides an international framework for establishing an Artificial Intelligence Management System (AIMS). Why, then, is another international standard necessary?

The answer lies in understanding the differences between organizational AI governance, conventional information security and threats specific to AI systems.

ISO/IEC 27090 does not replace ISO/IEC 42001. The two standards address different, complementary aspects of AI risk management.

However, understanding how they work together is more important than simply recognizing that both exist.


1. ISO/IEC 42001: Establishing accountable AI governance

Published in December 2023, ISO/IEC 42001 is the first international management-system standard dedicated to artificial intelligence.

It specifies requirements and provides guidance for establishing, implementing, maintaining and continually improving an Artificial Intelligence Management System.

Its primary objective is to help organizations manage AI responsibly through an established organizational framework.

That includes addressing AI-related risks and opportunities, assigning responsibilities, establishing policies and objectives, managing the AI lifecycle, assessing impacts and reviewing the effectiveness of the management system.

ISO/IEC 42001 follows the management-system approach found in other established international standards, using the Plan–Do–Check–Act cycle.

ISO/IEC 42001: The foundations of accountable AI governance and continual improvement

Its scope extends beyond cybersecurity.

It considers the wider organizational implications of developing, providing or using AI systems, including accountability, transparency, appropriate oversight and the potential effects of AI on affected parties.

Consider a company introducing an AI system to support important business decisions.

ISO/IEC 42001 provides a framework for asking:

  • Why is the organization using this AI system, and for what approved purpose?
  • Who is accountable for its deployment and operation?
  • What risks and impacts have been identified?
  • How are responsibilities shared with suppliers and other interested parties?
  • What forms of oversight, monitoring and review are appropriate?
  • How will the organization respond when the system changes or produces unexpected outcomes?

These are management and governance questions, although the answers can have significant cybersecurity implications.

Crucially, ISO/IEC 42001 already addresses risk, lifecycle management and the selection of appropriate controls. It should not be characterized as a standard concerned only with policy documents.

Its contribution is to establish an organizational system within which appropriate decisions, risk treatments and controls can be selected, implemented and evaluated.

ISO/IEC 42001 determines how an organization manages its AI responsibilities. It does not, by its existence alone, prove that every individual AI system is secure.

Source: ISO/IEC 42001:2023 – Official ISO publication

2. ISO/IEC 27090:2026: Understanding threats to AI systems

ISO/IEC 27090:2026 approaches the problem from a different perspective.

  • Its purpose is to help organizations understand security threats and compromises affecting AI systems throughout their lifecycle, including their consequences and approaches to detection and mitigation.
  • Its scope applies to organizations of different types and sizes that develop or use AI systems.

According to ISO’s official description, the new standard specifically addresses issues such as data poisoning and model theft while recognizing the continuing importance of established cybersecurity practices.

This distinction matters because AI systems introduce security considerations that cannot always be understood adequately through conventional infrastructure-focused assessments alone.

ISO/IEC 27090:2026 helps organizations understand, detect and mitigate AI-specific security threats across the AI lifecycle.

An organization may have strong identity management, endpoint security, network segmentation, logging and vulnerability-management practices, yet still face security threats involving the AI model, its data, its interactions or its updating process.

Consider the following examples:

Data poisoning

An attacker may attempt to manipulate data that a model uses during training or subsequent updates.

Depending on the model, the data pipeline and the attacker’s influence, this could change model behaviour, reduce reliability or introduce targeted errors.

The relevant security question is not simply whether the database is accessible to unauthorized users. It is whether the organization can establish sufficient trust in the origin, integrity and suitability of the information used to train or update the model.

Model theft or extraction

AI models may represent substantial intellectual property and commercial value.

Threats can involve unauthorized access to model artifacts or attempts to extract information through interactions with a deployed service.

Protection therefore involves considering both underlying access controls and the ways a model is exposed and used.

Adversarial manipulation

Some attacks are designed to influence how a model interprets inputs or produces outputs.

In machine-learning applications, adversarially manipulated inputs may affect classification or predictions. In generative AI applications, malicious instructions embedded in inputs or retrieved content may attempt to redirect model behaviour.

The mechanisms vary significantly across architectures and use cases. They should not be treated as one interchangeable category of attack.

Compromised retraining

One of the most consequential risks described in ISO’s public explanation concerns model updates.

Data collected during normal operation may eventually become part of a future training or retraining dataset.

If that information has been manipulated or corrupted, a compromise originating in operational data can influence later versions of the model.

This is especially relevant to organizations operating AI systems that evolve through periodic updates or continuous learning.

AI cybersecurity must consider not only where a model operates, but also what it learns from, what it trusts, how it is accessed and how it changes.

That is the specialized perspective ISO/IEC 27090 brings to the discussion.

Source: ISO/IEC 27090:2026 – Official ISO publication and IEC publication record.

3. The fundamental differences between the two standards

Although their subject matter overlaps, ISO/IEC 42001 and ISO/IEC 27090 are not equivalent documents.

ISO/IEC 42001 governs AI, while ISO/IEC 27090 strengthens its cybersecurity

The distinction can be summarized through two questions.

  1. ISO/IEC 42001 asks: Does the organization have an effective system for governing and managing its use or provision of AI?
  2. ISO/IEC 27090 asks: What AI-specific cybersecurity threats and compromises need to be understood, and how can their consequences be detected and mitigated?

These questions are related, but they are not interchangeable. A well-governed AI programme can still contain technical weaknesses.

Likewise, an AI application protected by sophisticated security controls can still be deployed without adequate accountability, impact assessment or operational oversight.

Effective AI assurance requires both perspectives.

4. Does ISO/IEC 27090 replace or complement ISO/IEC 42001?

The answer is clear: ISO/IEC 27090 complements ISO/IEC 42001. It does not supersede, amend or replace it.

ISO/IEC 27090 complements ISO/IEC 42001 by strengthening AI-specific cybersecurity within an accountable governance framework.

There is also an important third standard to consider: ISO/IEC 27001, which specifies requirements for an Information Security Management System (ISMS).

ISO explicitly explains that ISO/IEC 27090 complements the established security practices addressed by ISO/IEC 27001 and ISO/IEC 27002 by providing additional guidance on AI-specific threats.

This suggests a useful three-layer approach to organizational assurance:

ISO/IEC 42001 – AI governance

Establishes how the organization manages AI-related responsibilities, risks, controls and continual improvement.

ISO/IEC 27001 and ISO/IEC 27002 – Established information-security management and controls

Provide a framework for managing information-security risk and implementing appropriate security controls across relevant technology and information assets.

ISO/IEC 27090 – AI-specific cybersecurity guidance

Helps organizations examine threats and compromises involving AI models, training data, operational inputs and other AI-system components. These are complementary functions, not mutually exclusive alternatives.

An organization with an established ISMS and an AIMS may be able to incorporate AI-specific threat assessments and related safeguards into its existing governance, security and risk-management processes.

There is no inherent reason to establish an entirely separate management system simply because ISO/IEC 27090 has been published.

Equally, organizations do not need ISO/IEC 42001 certification before they can benefit from ISO/IEC 27090 guidance.

The practical objective is integration: one coherent approach to managing AI risks, with responsibilities assigned and technical safeguards selected according to the organization’s actual exposure.

5. A question for certified organizations: Is ISO/IEC 42001 enough?

This is where the distinction becomes particularly important for boards, CISOs and GRC professionals.

ISO/IEC 42001 supports independent management-system certification within an organization’s defined scope.

Certification can provide evidence that an organization’s AIMS has been assessed against the standard’s applicable requirements. However, certification should not be interpreted as a blanket security guarantee for every AI system developed or deployed by the organization.

An assessment of organizational management processes is not equivalent to comprehensive technical security testing of every application, model, integration and dataset.

ISO/IEC 42001 certification establishes AI governance, but effective AI cybersecurity requires additional technical assurance.

Consider an organization that has established its AI policy, documented relevant risk assessments, appointed responsible personnel and implemented appropriate oversight processes. These are important governance foundations.

But the organization must still consider whether its deployed AI services could be affected by manipulated datasets, exposed model interfaces, compromised dependencies or unauthorized model changes.

The two forms of assurance should reinforce each other.

ISO/IEC 27090 can help an organization make more informed security-risk decisions within its existing management arrangements.

This does not mean that every ISO/IEC 42001-certified organization necessarily has inadequate AI security. Nor does the publication of ISO/IEC 27090 invalidate existing certifications or automatically require a new certification audit. It means that security-risk assessments should remain current as relevant threats, technology and available guidance evolve.

The question should not be whether an organization holds an AI certificate. It should be whether its governance and security arrangements are effective against the risks of the AI systems it actually operates.

6. The EU AI Act: Where regulation enters the equation

The relationship between these standards becomes particularly relevant in the context of the EU AI Act.

Regulation (EU) 2024/1689 establishes a risk-based legal framework for artificial intelligence, with different obligations depending on the AI system, the relevant actor and the applicable provisions.

One of its key cybersecurity provisions is Article 15, which addresses accuracy, robustness and cybersecurity for high-risk AI systems. Article 15 requires high-risk AI systems to achieve appropriate levels of accuracy, robustness and cybersecurity throughout their lifecycle.

Its cybersecurity provisions specifically address resilience against attempts to alter the use, outputs or performance of AI systems by exploiting vulnerabilities.

They also contemplate measures addressing threats involving training data, pretrained components, adversarial inputs, confidentiality and model vulnerabilities.

The connection with ISO/IEC 27090 is therefore substantial.

ISO/IEC 42001 and ISO/IEC 27090 support EU AI Act compliance efforts through complementary AI governance and cybersecurity guidance.

The standard’s focus on AI-specific cybersecurity threats can help organizations understand and select measures relevant to the security outcomes addressed by Article 15.

ISO/IEC 42001, meanwhile, can help establish the wider organizational processes through which relevant AI obligations, responsibilities, risk assessments and controls are governed.

However, an important legal boundary must be respected.

Neither implementing ISO/IEC 42001 nor applying ISO/IEC 27090 automatically establishes compliance with the EU AI Act.

Under Article 40, a presumption of conformity depends on applicable harmonised standards whose references have been published in the Official Journal of the European Union, and only to the extent they cover the relevant legal requirements.

An international standard does not acquire that status merely because ISO and IEC have published it. Furthermore, a presumption of conformity under a harmonised standard should not be confused with ISO management-system certification.

The EU AI Act also includes separate provisions concerning conformity assessment and, under defined conditions, cybersecurity certification or compliance with relevant Cyber Resilience Act requirements.

These mechanisms have different legal foundations and cannot be collapsed into a general claim that an organization is “AI Act certified.”

As of October 2026, the EU AI Act is already partially applicable, while the amended framework provides later application dates for certain high-risk AI provisions:

  • 2 December 2027 for relevant Annex III high-risk systems
  • 2 August 2028 for relevant Annex I product-related systems.

The applicable obligations must be assessed for each system and role. Simply operating AI in shipping, logistics or critical infrastructure does not automatically make every application a high-risk AI system under the Act.

For decision-makers, the distinction is straightforward:

The EU AI Act establishes applicable legal obligations. ISO/IEC 42001 supports AI governance. ISO/IEC 27090 provides AI-specific cybersecurity guidance.

Each has value, but none can substitute automatically for another.

Sources: EU AI Act – Article 15, Article 40, Article 42, and European Commission implementation overview.

7. The numbers: What AI management-system certification tells us

The growth of AI governance standards is also becoming visible in international certification data.

The ISO Survey 2025, reporting accredited management-system certifications as of 31 December 2025, provides an early international baseline for ISO/IEC 42001.

Published analyses of the dataset identify:

ISO Survey 2025 reveals early ISO/IEC 42001 certification adoption, not the maturity or cybersecurity effectiveness of AI systems.
  • 271 recorded accredited ISO/IEC 42001 certificates.
  • 572 associated certified sites.
  • 40 countries and territories with recorded certificates.
  • 100 certificates in the United States, representing approximately 36.9% of the reported global total.

These figures require careful interpretation.

The ISO Survey counts certificates recorded within its defined accredited-certification dataset. It is not a count of all organizations using ISO/IEC 42001, all organizations with mature AI governance, or every AI application under management.

Coverage can vary by geography, certification body and reporting completeness.

The dataset also should not be used to calculate an unsupported growth rate from the preceding year, because ISO/IEC 42001 was newly included in the 2025 survey series.

Nevertheless, the figures provide a useful indication that formal AI management-system certification remains at an early stage compared with long-established management-system standards.

More importantly, the number of certificates tells us nothing directly about the effectiveness of AI-specific cybersecurity controls.

A certificate count cannot establish whether participating organizations have successfully defended their AI systems against model theft, data poisoning or other security compromises.

That is why governance metrics and technical assurance metrics must remain distinct.

Source: ISO Survey, with reported 2025 results cross-checked against published dataset analyses.

8. Why the distinction matters to maritime and critical infrastructure

The maritime industry illustrates why AI governance and AI cybersecurity cannot be treated as the same discipline.

Shipping companies, ship managers, port operators and maritime technology suppliers are exploring AI across business and operational activities.

AI governance and AI-specific cybersecurity work together to strengthen maritime operational resilience and manage safety-related risks.

Applications may include maintenance analytics, route planning support, document processing, anomaly detection, logistics optimization and cybersecurity operations.

The risks associated with these applications differ significantly.

Consider a hypothetical ship management company using an AI-powered predictive maintenance application. The system receives historical and operational machinery data, identifies patterns and generates recommendations for engineering teams.

From an ISO/IEC 42001 perspective, the organization needs to determine why the AI system is being used, who is accountable, what its limitations are, how its outputs are reviewed and how risks are monitored.

From an ISO/IEC 27090 perspective, it also needs to consider whether training or operational data can be manipulated, whether the model or its interfaces can be compromised, and whether model updates introduce new security exposures.

Now consider a hypothetical attempt to manipulate maintenance data. If the manipulated information reaches the AI application, it might affect the resulting recommendation.

Whether this leads to an operational consequence depends on the system architecture, the quality of validation, engineering oversight, the decisions made and the safeguards in place.

It would be technically incorrect to assume that compromising an AI application automatically means compromising vessel operational technology or machinery control.

That distinction matters.

AI-related cyber risk can influence operational decisions without an attacker ever obtaining direct control of onboard OT.

Conversely, a properly designed workflow with effective data validation, independent engineering review and controlled decision authority can help reduce the likelihood of an AI-generated error becoming an unsafe action.

This is where AI governance, security engineering and maritime operational resilience intersect.

The same reasoning applies to shore-based AI applications supporting port and terminal operations. The nature of the risk depends on the actual system connections, data dependencies, decisions and operational safeguards.

The objective is not to label every maritime AI system as safety-critical.

It is to understand which AI-supported decisions matter, what could undermine them, and what measures are proportionate to the consequences.

9. What should CISOs, boards and AI governance leaders do now?

The publication of ISO/IEC 27090 should trigger a focused review of existing AI-security risk assessments, not an automatic decision to establish another compliance programme.

The starting point is visibility.

Organizations need to know which AI systems are in use, including externally provided services and AI functions embedded in existing products.

They should establish the intended purpose, accountable owner, data dependencies, relevant suppliers, potential operational consequences and applicable legal requirements for each significant use case.

From there, the work should become more specific.

  1. For boards and executive management, the priority is obtaining credible assurance that consequential AI risks are understood, assigned and managed. Reporting should distinguish organizational control maturity from technical security evidence.
  2. For CISOs and security teams, the priority is assessing AI-specific attack surfaces and determining whether existing information-security controls adequately address model, data, interface and lifecycle risks.
  3. For AI governance and GRC leaders, the priority is connecting the results of technical assessments to organizational risk ownership, treatment decisions, oversight processes and management review.
  4. For AI engineering and application owners, the priority is understanding how systems behave under adverse conditions, how updates are controlled, how model and data integrity are assessed, and how security issues are detected and addressed.
  5. For maritime technical and operational leaders, the priority is ensuring that AI-supported decisions have appropriate operational validation, human oversight and fallback arrangements, especially where errors could influence safety or continuity.

Rather than duplicating existing processes, organizations should connect AI-specific security findings to their established risk registers, change-management arrangements, incident-response procedures, supplier oversight and internal assurance activities.

The necessary evidence will depend on the AI application. It may include threat assessments, model and data provenance records, security-testing results, access-control reviews, monitoring records, change approvals and documented operational exercises.

What matters is whether the organization can demonstrate that selected measures are appropriate and effective for its actual risks.

10. The bigger lesson: Governance does not equal security, and security does not equal governance

The publication of ISO/IEC 27090:2026 reinforces an important principle.

AI risk cannot be managed adequately through documentation alone, nor can it be reduced to a collection of technical safeguards.

An organization can establish sophisticated policies, roles and risk processes while still overlooking important vulnerabilities in its AI applications.

Another organization may deploy strong technical security measures while lacking sufficient accountability over what AI systems are permitted to do, how their recommendations are used or how their wider impacts are assessed.

Both are weaknesses.

AI governance and AI-specific cybersecurity are complementary disciplines, both essential for responsible, secure and resilient AI adoption.
  • ISO/IEC 42001 provides a management-system framework for governing AI.
  • ISO/IEC 27090 contributes specialized understanding of AI-specific cybersecurity threats and their potential consequences.
  • ISO/IEC 27001 and ISO/IEC 27002 remain relevant to established information-security management and controls.

Together, these references can support a more coherent approach to managing AI-related risks, provided organizations translate them into actual responsibilities, safeguards, testing and evidence.

The arrival of ISO/IEC 27090 does not mean that organizations should replace their AI governance frameworks. It means that their governance must be informed by an increasingly mature understanding of how AI systems can be attacked and compromised.

For executives, CISOs and operational decision-makers, that is the distinction worth remembering.

The objective is not simply to demonstrate that AI is governed. It is to demonstrate that AI is governed responsibly, protected proportionately and used within operationally acceptable limits.


About the author

Alexandros Engelen is an independent cybersecurity strategist specialising in maritime cyber risk. His work examines the relationship between cybersecurity threats, technology governance, operational resilience, regulation and the real-world consequences of digital dependencies across the maritime industry.


References

  1. ISO – ISO/IEC 42001:2023: Artificial intelligence management systems.
  2. ISO – ISO/IEC 27090:2026: Addressing security threats and compromises to AI systems.
  3. IEC – ISO/IEC 27090:2026 official publication record.
  4. ISO – ISO/IEC 27001:2022.
  5. ISO – The ISO Survey.
  6. European Union – Regulation (EU) 2024/1689, EU AI Act.
  7. European Commission – Article 15: Accuracy, robustness and cybersecurity.
  8. European Commission – Article 40: Harmonised standards.
  9. European Commission – Article 42: Presumption of conformity with certain requirements.
  10. European Commission – AI Act implementation and timeline.

#ISO42001 #ISO27090 #AIGovernance #AICybersecurity #EUAIAct #GRC #CyberRiskManagement #MaritimeCyberSecurity #Governance


Editorial methodology:

  • This analysis uses official publication descriptions and publicly accessible regulatory material, supplemented by certification-data reporting.
  • It does not reproduce the full licensed texts of the standards or purport to provide an exhaustive clause-by-clause mapping.
  • Maritime examples are hypothetical and do not describe disclosed incidents.
  • Regulatory discussion is informational and is not a determination of legal compliance for any particular organization.

A Verified Identity, Not Verified Intent: How PoisonX Blinds a Control Maritime Cyber Compliance Requires

Maritime Cyber Threat Briefing #13

Article content
A Trusted Signature Proves Origin—Not Intent

Briefing #12 examined what happens when the network perimeter fails. When an edge device that was supposed to keep an adversary out becomes the entry point instead. This briefing examines the layer behind it: the endpoint security that is supposed to notice an adversary once they are already inside. A ransomware operation documented by Symantec and Carbon Black in July 2026 defeats that layer using a mechanism that should not exist. It does not exploit a flaw in endpoint software. It carries a genuine Microsoft signature, obtained through the process Microsoft itself designed to establish trust in kernel-mode code, and uses that trust to make the security software blind rather than absent. The agent keeps running. It simply stops seeing.

No maritime or transportation-sector victim of this campaign has been publicly confirmed. This briefing is an exposure analysis, not incident reporting, and it should be read as such throughout. The reason it belongs in this series regardless is structural: the control this technique defeats is not optional in modern vessel cyber-resilience design. It is one IACS explicitly requires shipowners to install.

The ransomware, precisely

The threat actor tracked by Symantec as Hyadina has run one continuous ransomware-as-a-service operation since March 2022, rebranding twice while keeping the same core tooling and operational habits. Monster appeared first, a Delphi-based locker targeting 32-bit Windows systems. In June 2024 the operation rebranded as Beast, adding Linux and VMware ESXi support and multilingual capability. In May 2026, Symantec identified the third iteration: GodDamn.

In the intrusion Symantec and Carbon Black investigated, the attackers gained initial access by a vector that could not be determined [account compromise is the common assumption for this actor’s profile] then established persistence with AnyDesk, installed in a user’s Music folder on the first host as a manual, hands-on-keyboard action rather than an automated deployment. Credential harvesting followed, using Mimikatz alongside a NirSoft-based toolkit, with PsExec for lateral movement. On a second host, the attackers staged a binary named symantec.exe, disguised as the defender’s own product [a small but telling detail], since it is designed to look normal to an administrator glancing at a process list rather than to defeat automated detection. That binary dropped the kernel driver at the centre of this briefing.

Five days elapsed between initial access and ransomware detonation, across at least ten hosts, before encryption began.

PoisonX and the mechanism that matters

Article content
Trusted by the System. Weaponized Against Its Defenses

The driver, filename g11.sys, carries a signature attributed to Microsoft Windows Hardware Compatibility Publisher. Once loaded, it operates at kernel privilege to terminate security-product processes and strip the user-mode API hooks that antivirus and EDR agents rely on to observe system activity. Symantec’s assessment is direct: the driver has no legitimate use case, and its adoption by Hyadina represents a deliberate escalation in defensive-evasion capability.

This is commonly described in press coverage as a Bring-Your-Own-Vulnerable-Driver, or BYOVD, attack, and the label is close enough to be useful but not precise enough to guide defense. Classic BYOVD abuses a flaw in an existing legitimate driver. A vulnerability the attacker did not create but discovered. PoisonX appears to have been built from the outset for this purpose and still obtained a valid signature, which is a different and in some ways more serious problem, because there is no underlying software flaw to eventually patch.

How that signature was obtained is worth understanding precisely, because the explanation is not “Microsoft was fooled”. To submit a driver for attestation signing under the Windows Hardware Compatibility Program, a developer needs a Hardware Dev Center account backed by an Extended Validation code-signing certificate. A credential that requires real-world identity and organizational verification to obtain. Attestation signing, the relevant path for this kind of submission, does not require the formal hardware lab certification testing that full WHCP certification does. Microsoft’s own documentation states plainly that attestation-signed drivers are trusted by Windows, but their functionality is not guaranteed. The programme verifies who is submitting the code. It does not verify what the code does. Attestation signing establishes publisher identity and package integrity, but it is not equivalent to full behavioural or compatibility certification. Microsoft’s own documentation states that attestation-signed drivers do not undergo Hardware Lab Kit testing, are not Windows Certified, and carry no assurance regarding compatibility or functionality. That distinction is central to the PoisonX case: a valid signature can tell Windows who submitted a package and whether it was altered after signing, but it does not, by itself, prove that the driver’s operational intent is benign. Signing Options

The failure is therefore not that digital signing provides no value. It is that identity assurance and “software behavior” assurance are different controls. And the industry too often treats the first as though it guarantees the second.

The Chain of Inherited Trust

Article content
When Trust Propagates Unchallenged, Assurance Becomes the Blind Spot

The deeper issue is not a single signing decision. It is the chain of trust that follows it.

Windows accepts the driver because its package carries a valid Microsoft signature. Endpoint software operates on the assumption that permitted kernel code belongs inside the trusted computing base. Asset-management systems then record that the security agent remains installed and running. Audit and compliance processes may accept that recorded state as evidence that malware protection is operational.

Each layer inherits confidence from the one before it. PoisonX exploits that inheritance. It does not need every layer to fail independently. It needs the first trust decision to propagate through the rest of the environment.

The signature validates package provenance and integrity. The operating system converts that into permission. Monitoring converts permission into apparent normality. Compliance converts apparent normality into assurance.

This is why the incident matters beyond Windows driver signing. It demonstrates how a narrow technical trust decision can become an inaccurate board-level statement about operational resilience.

PoisonX Is a Capability, Not a Campaign

Article content
One Capability. Many Campaigns. The Same Defensive Blind Spot

PoisonX did not appear for the first time inside the GodDamn intrusion. The driver had already been documented earlier in 2026 in a separate campaign targeting organizations in Japan and China, where it was deployed alongside the modular 10FXRAT remote-access tool. Symantec later observed it in a financially motivated GodDamn ransomware operation. The available reporting does not establish that the two campaigns were conducted by the same actor, and they should not be conflated. What it does establish is more strategically important: the same kernel-level defensive-evasion capability has moved across distinct operational contexts. GodDamn Ransomware

That makes PoisonX better understood as an access-enablement component than as a feature belonging exclusively to one ransomware family. Once a signed driver capable of degrading endpoint visibility enters circulation, it can be reused, transferred, sold, or incorporated into different intrusion chains. The defender is therefore not dealing only with Hyadina. The defender is dealing with the availability of a reusable capability whose value applies equally to espionage, credential theft, ransomware, and long-term persistence.

For maritime organisations, this distinction matters. Threat assessments built around named ransomware groups can become obsolete when techniques and components migrate faster than actor labels. The relevant question is not only whether GodDamn targets shipping. It is whether any actor targeting shipping can obtain or reuse the same driver.

A Recurring Failure in the Trust Model

This is not a novel failure. In December 2022, following joint reporting from Mandiant (part of Google Cloud), SentinelOne and Sophos, Microsoft revoked certificates associated with maliciously signed drivers in the Poortry and Stonestop case. Those drivers followed a closely related pattern: trusted kernel code built or repurposed to disable security software and support ransomware operations. GodDamn and PoisonX therefore represent the recurrence of a structural weakness in the trust model, not the appearance of an entirely new one.

Four years after a mass certificate revocation addressed the previous wave, the same trust-model problem has returned in a new operational form. The lesson is that revocation removes known keys: it does not redesign the system that issued trust.

Why the two standard defenses are weaker in practice than the coverage implies

Microsoft’s own guidance for this threat class centre on two controls, and both deserve a more precise treatment than “keep them updated”.

  1. The Microsoft Vulnerable Driver Blocklist is an important control, but it is not a complete answer to newly weaponized signed drivers. Microsoft currently updates the blocklist quarterly and also distributes blocklist changes through monthly Windows servicing. Organisations can additionally deploy the latest recommended rules through App Control for Business. Even so, newly identified drivers may remain usable before they are evaluated, added, distributed, and enforced across every affected system. Microsoft also acknowledges that some blocks may be delayed where immediate enforcement could break legitimate devices or software. The practical gap is therefore no longer best described as an annual update problem. It is the interval between discovery, policy inclusion, fleet-wide distribution, enforcement, and operational validation. Microsoft Learn
  2. Hypervisor-Protected Code Integrity (HVCI, also called Memory Integrity) availability and suitability depend on a combination of processor, firmware, virtualization, Secure Boot, IOMMU, and driver-compatibility conditions. Microsoft’s current automatic-enablement baseline generally centre on Intel eighth-generation or later processors and “AMD Zen 2” or later, together with supporting UEFI and platform-security features. Older systems may fail to meet that baseline, may lack required capabilities, or may incur unacceptable compatibility and performance consequences. In maritime environments, the decisive question is therefore not simply whether HVCI exists in Windows, but whether the specific OEM-controlled hardware and software configuration can support and enforce it without affecting an approved operational system. Memory Integrity

Maritime exposure: what is documented and what is not

To be precise about the limits of what can be asserted here: Hyadina’s confirmed victim profile across all three ransomware generations spans

  • manufacturing
  • construction
  • healthcare
  • business services
  • education.

No maritime, port, or logistics victim has been publicly reported. The group’s targeting pattern: opportunistic, financially motivated, avoiding machines geolocated in the Commonwealth of Independent States. This shows no evidence of deliberate maritime interest. This briefing is not reporting a maritime attack. It is describing why the maritime sector is structurally exposed to this class of technique regardless of who deploys it next.

That exposure rests on three points, each independently documented.

  1. The fleet’s underlying Windows estate skews toward exactly the hardware generation this control gap concerns. Marlink‘s own fleet telemetry, published around the October 2025 end-of-support date for Windows 10, found that just over forty percent of surveyed maritime systems were still running Windows 10, against roughly half on Windows 11. A meaningful share of that population sits on hardware or vendor-locked images that predate HVCI’s hardware baseline.
  2. The upgrade path on much of this equipment does not run through the ship’s IT department. Bridewell‘s analysis of the maritime Windows 10 exposure makes a specific and important point: an ECDIS terminal or other OEM-locked navigation console built on a fixed vendor image frequently has no viable upgrade path without manufacturer recertification. A shore-based IT team cannot simply enable Memory Integrity on a bridge system the way it could on an office laptop, even where the underlying hardware would technically support it. Because doing so may invalidate the vendor’s type approval or void support arrangements. Remediation runs through the equipment manufacturer’s release cycle, not the ship’s patch cycle. And that cycle, for many maritime OEMs, is measured in years rather than months.
  3. The specific control PoisonX defeats is not discretionary under current classification requirements. IACS UR E26 states plainly that malware protection shall be implemented on computer-based systems within its scope. That where an industry-standard anti-virus or anti-malware product is available for the operating system in question, it shall be installed, maintained and kept regularly updated, subject to a narrow exception where installation would impair the system’s function. PoisonX does not need to defeat a theoretical or aspirational control. It defeats the exact control class E26 obligates vessels contracted after 1 July 2024 to run. It does so in a way that leaves the required software installed, apparently functioning, and silently blind. A vessel may appear compliant through conventional evidence [the product is installed, licensed, updated, and reporting a healthy status] while a PoisonX-class driver has degraded the underlying visibility on which that status depends.

Detection: what to look for, without a working exploit as a guide

Article content
When the Agent Goes Quiet, the Silence Becomes the Signal

Consistent with this series’ standard on attack methodology, the guidance below stays at the level of observable indicators and architectural response, not driver-loading technique.

  • Unexplained new kernel drivers, regardless of signature status. A signed driver is not a trusted driver by default. It is a driver whose submitter was identity-verified. Any newly appearing “*.sys” file that was not part of a known, change-controlled installation [vendor updates, planned maintenance software, approved monitoring agents] warrants investigation on that basis alone. Independent of whether it carries a valid Microsoft signature.
  • Security software that goes quiet rather than absent. The specific danger of this technique is that the endpoint agent’s process continues to run and continues to appear “installed” in inventory tools, while its actual visibility into system activity has been severed. A monitoring regime built only around “is the agent running” will not catch this. Regimes that also verify the agent is actively generating telemetry, heartbeat and event-volume monitoring, not just process-presence checks: Will.
  • Impersonation of security-product filenames or installers. The “symantec.exe” staging file in the documented intrusion was named specifically to look unremarkable to an administrator scanning a process list. File-name matching alone is a weak control, but combined with hash and publisher verification against known-good vendor releases, it closes an easy gap.
  • Cross-reference published indicators regardless of whether the local EDR has flagged anything. Symantec has published the SHA-256 hashes for PoisonX and the associated staging binary in its Protection Bulletin. The entire premise of this technique is that a blinded EDR will not flag what happens after it is blinded. So the indicator sweeps that do not depend on the local agent having caught the initial compromise are the only reliable way to confirm a system is clean.

Look for disagreement between systems

A blinded endpoint often reveals itself through disagreement rather than absence. Identity systems may record privileged logons that the endpoint does not explain. Network sensors may observe east-west connections or remote administration traffic while the host produces no corresponding process telemetry. A central log platform may show a sudden fall in event volume from one machine even though the device remains online and the agent continues to report a healthy status.

The detection model should therefore compare independent sources:

  • endpoint event volume and heartbeat;
  • identity and privileged-logon activity;
  • network and DNS telemetry;
  • firewall, VPN, and remote-access records;
  • service and driver installation events forwarded off-host.

When the network says activity occurred and the endpoint says nothing happened, the silence is itself an indicator.

Mitigation, calibrated to what a maritime IT/OT estate can actually do

Article content
Real Maritime Resilience Begins With Controls the Estate Can Actually Sustain
  • Where HVCI is achievable, enable it, and treat the exercise as a compatibility project rather than a toggle. For shore-side infrastructure and any bridge or engine-room Windows system on hardware from the qualifying generation onward, Memory Integrity meaningfully raises the cost of this technique. Budget for driver-compatibility testing before enabling it in production, since incompatible third-party drivers will be blocked and can affect specialised maritime software.
  • Where HVCI is not achievable [which will be a substantial share of the fleet for some years yet] compensate architecturally. This is precisely the case IACS UR E26 network segmentation exists to address: isolate the systems that cannot receive this class of protection, restrict what can reach them, and ensure that a compromise on an unprotected navigation or cargo system cannot pivot laterally into segments that can be protected. A vessel’s cyber-resilience posture should assume that some endpoints will remain permanently unable to run the strongest available kernel protections, and design the network around that reality rather than around an assumption that every system will eventually be upgraded.
  • Treat “the antivirus is installed and shows green” as a necessary but insufficient compliance signal. E26 compliance audits and internal cyber risk assessments should distinguish between confirming a product is installed and confirming it is “actively generating telemetry”. The gap between those two states is exactly where this technique lives.
  • Restrict driver installation privileges as a baseline control, independent of signature status. Loading a new kernel driver should require the same change-control discipline as any other privileged system modification, on shore infrastructure and on vessel IT systems alike.
  • Maintain offline, verified indicator feeds for vessels with limited connectivity. Given that vessels cannot always pull real-time threat intelligence, shore-based cyber teams should periodically push verified indicator sets [including published hashes for campaigns like this one] to fleet systems during scheduled connectivity windows. Rather than relying solely on the vulnerable driver blocklist’s own update cadence, which is already slow even for well-connected shore networks.

Five questions management should ask

  1. Can we identify every kernel driver authorised on critical maritime systems?
  2. Can we detect when endpoint telemetry degrades while the agent remains online?
  3. Which vessel systems cannot support HVCI or equivalent protection and why?
  4. Who is responsible for remediating an OEM driver-control limitation?
  5. Can we independently demonstrate that malware protection is effective, not merely installed?

Regulatory considerations

Article content
Compliance Is Only Proven When the Control Still Works
  • Mandatory. IACS UR E26 and E27, applicable to vessels contracted for construction on or after 1 July 2024, are the direct regulatory anchor for this briefing. E26’s explicit requirement for installed and maintained malware protection is the control this technique targets, and E26’s network segmentation requirements are the primary compensating control where endpoint protection cannot be fully realized. The International Maritime Organization Guidelines on Maritime Cyber Risk Management, referenced against MSC.428(98) and discussed at length in my Maritime Cyber Insights Article #4, establish the risk-management framework within which this exposure should be assessed and documented. Though they do not specify technical controls at this level of detail. Where NIS2 applies through the relevant national implementing legislation, its requirements for cybersecurity risk management, incident detection, response, governance, and reporting are directly relevant to the distinction between a security product being installed and the underlying security capability remaining operational. Applicability and implementation must still be assessed jurisdiction by jurisdiction. As of July 2026, most EU Member States had notified transposition measures, but Ireland, Spain, France, and the Netherlands had not notified full transposition and were referred by the European Commission to the Court of Justice of the European Union.
  • Not yet mandatory, but directly relevant. Neither the IMO framework nor IACS E26/E27 currently mandates a specific standard for kernel-level driver control (HVCI or equivalent) or for hardware baselines capable of supporting it. This is a genuine gap between what current regulation requires [malware protection software be present] and what a sophisticated adversary has demonstrated defeats that requirement without violating it.

PoisonX also exposes a responsibility gap between the shipowner, the system integrator, the OEM [Original Equipment Manufacturer], and the classification process. A shipowner may be required to maintain malware protection while lacking authority to modify the underlying driver policy, operating-system image, or security configuration of an OEM-controlled system. The owner carries the operational risk, while the technical ability to harden the platform may sit with the manufacturer.

Future assurance should therefore ask OEMs to provide more than a statement that antivirus is supported. For each Windows-based system, owners and class should seek:

  • the supported kernel-security configuration
  • HVCI and driver-control compatibility status
  • an approved-driver inventory
  • the process and maximum timeline for revoking a compromised driver
  • a defined telemetry-health test
  • compensating controls where platform limitations prevent modern enforcement

This turns malware protection from a product checkbox into an evidenced system capability.

Article content

The uncomfortable conclusion of this briefing is not that GodDamn will target a vessel next. It may never do so, its operators show no sign of interest in the sector. The conclusion is that the trust mechanism PoisonX exploits is a general-purpose weakness available to any actor able to obtain or reuse a maliciously signed driver, and that maritime cyber assurance places significant reliance on the exact control class this technique can silently degrade.

The real lesson of PoisonX is not that signatures have become meaningless. They have not. Signatures establish provenance and integrity, and both remain essential. The lesson is that provenance is not purpose, integrity is not safety, and installation is not effectiveness.

Maritime cyber compliance increasingly requires controls that sophisticated actors know how to leave visibly present and operationally useless. The next maturity step is therefore not another checkbox. It is the ability to prove [continuously and independently] that the required control is still performing the function for which it was installed.

A verified identity is not a verified intent. A running agent is not a seeing agent. A compliant system is not necessarily a resilient one.


Maritime Cyber Threat Briefing is an independent series covering cyber threats, vulnerabilities, and risk management across the global maritime industry. It is published by Alexandros Engelen, a Cybersecurity Strategist, specializing in maritime cyber risk.

Sourcing note. This briefing draws on primary and authoritative sources including the Symantec and Carbon Black (Broadcom) Threat Hunter Team’s published analysis and Protection Bulletin covering the GodDamn/PoisonX intrusion; Microsoft Learn documentation on driver code-signing requirements, attestation signing, the Windows Hardware Compatibility Program, and Hypervisor-Protected Code Integrity hardware requirements; Microsoft’s own support documentation on the Vulnerable Driver Blocklist update cadence; contemporaneous reporting on the December 2022 Poortry/Stonestop certificate revocations; Marlink’s published fleet telemetry on maritime Windows 10 prevalence; Bridewell’s analysis of maritime OEM-locked system upgrade constraints; and the primary text of IACS Unified Requirement E26. Attribution of GodDamn to the Hyadina threat actor and its lineage through Monster and Beast follows Symantec’s published assessment. No maritime, port, or logistics-sector victim of this campaign has been publicly confirmed at the time of writing; all maritime-specific content in this briefing is exposure analysis based on documented fleet characteristics and regulatory requirements, not incident reporting. Discussion of the driver’s mechanism is deliberately confined to the level required for detection and defence; no driver-loading technique, exploitation detail, or code-signing circumvention method beyond what Microsoft itself has published is included. No commercial relationship exists between the author and any vendor, product or service referenced.

What are IMO Resolution MSC.428(98) and the IMO Guidelines on Maritime Cyber Risk Management?

Maritime Cyber Insights #4

Article content
Cyber resilience begins beyond compliance.

Many shipping companies state that they “comply with MSC.428(98).” That shorthand is understandable, but technically incomplete.

International Maritime Organization Resolution MSC.428(98) is not a standalone company certification standard, nor did it establish a separate maritime cyber certificate. The mandatory framework is found in SOLAS chapter IX and the International Safety Management Code, as implemented by the relevant flag Administration. MSC.428(98) confirms that cyber risk management must be addressed within an approved Safety Management System in accordance with the objectives and functional requirements of the ISM Code.

This distinction is more than semantic.

It determines how the obligation is implemented, where cyber risk should appear within the Safety Management System, what evidence auditors and inspectors may expect, how deficiencies may be recorded and how failure to implement the company’s own procedures can escalate into an ISM finding.

It also explains why a company can hold a valid Document of Compliance while still remaining materially exposed to cyber risk. Certification confirms that a management system has been established and verified at particular points in time. It does not guarantee that every procedure is consistently followed across every vessel, department, supplier and operational environment.

This article explains what MSC.428(98) actually says, what the current IMO Guidelines require following their substantial 2025 revision and subsequent 2026 update, and how the framework relates to the ISM Code, IACS Unified Requirements E26 and E27, national or regional legislation and industry guidance.

It is written for the people who must operationalize these requirements, not merely describe them in a compliance statement.

What MSC.428(98) actually says

Resolution MSC.428(98), Maritime Cyber Risk Management in Safety Management Systems, was adopted by the IMO Maritime Safety Committee on 16 June 2017.

  • Its operative content is brief.
  • It affirms that an approved Safety Management System should take cyber risk management into account in accordance with the objectives and functional requirements of the ISM Code.
  • It encourages Administrations to ensure that cyber risks are appropriately addressed within Safety Management Systems no later than the first annual verification of the company’s Document of Compliance after “1 January 2021”.
  • It acknowledges that precautions may be necessary to preserve the confidentiality of certain aspects of cyber risk management.
  • It requests Member States to bring the resolution to the attention of relevant maritime stakeholders.

It did not establish a separate certification regime

MSC.428(98) did not amend the ISM Code or create a separate cybersecurity certification process. Instead, it removed any remaining ambiguity about whether the ISM Code’s existing risk-management obligations extend to cyber risk.

The ISM Code requires companies to assess all identified risks to their ships, personnel and the environment and to establish appropriate safeguards. MSC.428(98) confirms that cyber threats, vulnerabilities and technology-related operational failures fall within that risk-management framework.

The practical obligation therefore rests within the approved SMS. Not in a separate declaration of “MSC.428 compliance.”

Its implementation call is primarily directed at Administrations

The resolution’s direct implementation language primarily addresses flag Administrations.

Its effect reaches shipping companies and vessels through flag-State implementation, company DOC verification, shipboard SMC verification, recognized-organization audits and, where applicable, Port State Control.

This is why deficiencies normally arise under the ISM Code, the company’s approved SMS or another applicable legal requirement. They are not usually treated as findings against a standalone “MSC.428 standard.”

The implementation deadline has passed

The first annual DOC verification after 1 January 2021 has long since taken place for affected companies, and several annual verification cycles have now followed.

The relevant question in 2026 is therefore no longer whether the company has inserted a cybersecurity section into its SMS.

The real question is whether cyber risk management has been effectively integrated into normal company and vessel operations—and whether the company can demonstrate that integration through objective evidence.

That is a materially higher standard than possessing a policy document.

The Guidelines: the 2025 overhaul and the 2026 update

Article content
Rev.4 is the current benchmark; Rev.3 delivered the transformation.

MSC.428(98) establishes the regulatory direction. The operational detail is contained in the jointly approved IMO Guidelines on Maritime Cyber Risk Management.

The current instrument is MSC-FAL.1/Circ.3/Rev.4, issued on 28 May 2026 following approval by the Facilitation Committee at FAL 50 in March 2026 and the Maritime Safety Committee at MSC 111 in May 2026.

Rev.4 carries forward the substantial changes introduced through Rev.3 in April 2025 and adds the IAPH Cyber Resilience Guidelines for Emerging Technologies in the Maritime Supply Chain to the non-exhaustive list of supporting industry guidance in section 4.3.

The principal substantive transformation therefore occurred in Rev.3 and remains fully reflected in Rev.4.

A new terminology baseline

The current Guidelines are built around the concept of the Computer Based System (CBS).

A CBS is defined as a programmable electronic device [or an interoperable set of such device] organized to collect, process, maintain, use, share, disseminate or dispose of information.

Onboard CBSs include both information technology and operational technology systems. They may consist of connected subsystems and may communicate with shore-based systems, other ships or external facilities through private or public communications infrastructure.

This terminology is important because it creates closer alignment between the IMO Guidelines and the language used in IACS UR E26, UR E27 and UR E22.

It also discourages the artificial separation of “IT cybersecurity” from the operational systems on which navigation, propulsion, power generation, cargo operations, communications and vessel safety increasingly depend.

A sixth functional element: Govern

Earlier versions of the Guidelines followed five familiar functional elements:

  1. Identify
  2. Protect
  3. Detect
  4. Respond
  5. Recover.

The revised framework adds Govern at the beginning, reflecting the governance function introduced through the NIST Cybersecurity Framework 2.0.

Under Govern, companies should establish and monitor cyber risk-management strategies, expectations and policies, define roles and responsibilities. And ensure business continuity, backup management, disaster recovery and crisis management.

The Guidelines also state that a person or entity should be designated as accountable for the planning, resourcing and execution of cybersecurity activities. That person or entity should be given the authority, support, knowledge and expertise required to perform the role.

This establishes a clear IMO governance expectation.

Cyber risk should not be nominally assigned to an IT department that has no authority over vessel operations, procurement, crewing, third-party access or operational technology. Nor should it be assigned to the Designated Person Ashore purely because the DPA already occupies a recognized position within the ISM structure, unless the DPA is given the necessary cybersecurity competence, authority, support and resources.

Accountability without authority is not governance.

Controls described as minimums

The current Guidelines continue to describe themselves as providing high-level recommendations. They remain a non-mandatory IMO circular and predominantly use the term “should.”

However, paragraph 3.5 states that the functional and technical cybersecurity controls listed under each functional element represent the minimum controls that should be implemented, with additional controls considered according to the identified cyber risks.

This wording does not convert the circular into a directly mandatory instrument.

An auditor or inspector should still ground a formal finding in an applicable ISM requirement, flag-State requirement, class rule, national regulation or provision of the company’s approved SMS. Not simply in the failure to follow a non-mandatory circular.

Nevertheless, the language materially strengthens the Guidelines’ practical value.

They now provide an IMO-issued benchmark against which the adequacy of a company’s cyber risk-management arrangements may be evaluated. The scope for claiming that almost any minimal arrangement is sufficient has narrowed considerably.

Named functional and technical controls

The Guidelines now expressly address measures including:

  • Unique user credentials and the separation of standard and privileged accounts.
  • Deactivation of accounts belonging to departing personnel.
  • Removal of default passwords and implementation of strong password policies.
  • Multi-factor or continuous authentication where appropriate.
  • Limitation of exploitable internet-facing services.
  • Hardware and software approval processes.
  • Secure collection and retention of logs for intrusion detection and incident response.
  • Segmentation of OT device networks from IT networks.
  • Security controls for systems connected to the internet, corporate intranets, third parties and ship-to-port interfaces.
  • Policies governing the use of cryptography.
  • Controls against unauthorized removable media.
  • Regular system backups and software updates.
  • Maintained incident-response plans.
  • Software and hardware supply-chain policies for systems identified as critical.
  • Periodic assessment, auditing, review and updating of cyber risk-management measures.

The Guidelines also make clear that effective risk assessment should extend beyond systems physically installed onboard..

Cyber risk assessment should consider third-party vendors, embedded systems, hardware and software supply chains, maintenance devices, remote access arrangements and the interdependencies between safety-critical systems. That is essential because a vessel’s cyber exposure does not end at the ship’s physical boundary.

The human element

The Guidelines call for annual basic cybersecurity training for all employees, OT-specific cybersecurity training for OT users and cybersecurity familiarization for crew members when they join the vessel.

  • Training should address cyber hygiene, recognition and detection of an ongoing cyber incident, response and recovery.
  • Cybersecurity knowledge should also be tested periodically. For example, through drills and exercises. The document does not state that cyber drills themselves must be conducted annually, and that distinction should be preserved.
  • The “Respond and Recover” elements also expect cyber incidents to be recorded, reported to the necessary parties within timeframes defined by the relevant Administration and subjected to root-cause analysis to resolve underlying issues and reduce the likelihood of recurrence.
  • Documents created to satisfy the functional elements should themselves be protected against unauthorized access, deletion, destruction or amendment.

The training, drill, reporting, record-keeping, corrective-action and root-cause language is not incidental. It reflects the vocabulary and management logic of the ISM Code.

Integration into the Safety Management System

Article content
Cyber risk must be embedded across the entire ISM framework.

The IMO position is that cyber risk management should be incorporated into existing risk-management and safety-management processes. It should not operate as a disconnected parallel management system.

A standalone cybersecurity manual may still be useful as a controlled supporting document, but only where its requirements are linked to the relevant SMS procedures, responsibilities, records, audit processes and operational controls.

Otherwise, it becomes a compliance attachment rather than part of the way the company manages safety.

Proper integration should be visible across the ISM architecture.

ISM 1.2.2.2: Risk assessment and safeguards

Cyber threats and vulnerabilities should enter the same structured risk-assessment process used for other safety and operational hazards.

The assessment should consider the vessel’s type, operating profile, trading pattern, onboard-system complexity, external connectivity, data dependencies, remote support arrangements and potential consequences of compromised availability, integrity or confidentiality.

It should be reviewed after significant system modifications, network changes, new vendor connections, cyber incidents, changes in operational profile or the identification of material new vulnerabilities.

ISM 3: Company responsibilities and authority

The company should define who is accountable for cyber risk management and how authority is distributed between management, IT, OT specialists, marine operations, technical departments, procurement, crewing, safety functions and vessels.

Responsibility should be documented, but it must also be operationally credible. The person held accountable must be capable of obtaining information, directing action, escalating unresolved risks and securing the necessary resources.

ISM 5: Master’s responsibility and authority

The Master should understand the vessel’s cyber-related operational risks and the procedures that apply when system integrity, availability or trustworthiness is in doubt.

The SMS should preserve the Master’s overriding authority to take decisions necessary for safety, including suspending unsafe digital processes, limiting external connectivity or reverting to manual or degraded operating modes where appropriate.

ISM 6: Resources and personnel

Competence requirements, familiarization, shore-based support and recurrent training should include an explicit cyber dimension.

General cybersecurity awareness alone is insufficient for personnel who operate, maintain or administer safety-critical OT systems. Training should reflect actual responsibilities.

Bridge personnel, engineers, electro-technical officers, IT support teams, procurement staff, superintendents and senior management do not all require the same level or type of cyber competence.

ISM 7: Shipboard operations

Procedures for key shipboard operations should consider the CBSs on which those operations depend.

This includes navigation, propulsion, steering, power generation, cargo management, ballast operations, communications, safety systems and other functions whose disruption or manipulation could create a hazardous situation.

Procedures should address what happens when a system becomes unavailable, produces unreliable data or may have been compromised.

ISM 8: Emergency preparedness

Cyber incidents should be incorporated into emergency planning and exercises.

Scenarios may include loss of navigation data, ransomware affecting shore or vessel systems, compromised remote access, loss of communications, failure of cargo-management systems, malicious alteration of operational data or simultaneous loss of multiple interconnected systems.

The company should define technical and operational decision-making responsibilities, communication arrangements, escalation paths and procedures for safe degraded operation.

ISM 9: Reporting and analysis

Cyber incidents and relevant near-misses should enter the established reporting, investigation and corrective-action process. The objective should not be limited to restoring a system.

The company should determine why the incident occurred, which procedural or technical controls failed, whether similar exposure exists elsewhere in the fleet and what corrective or preventive measures are required.

ISM 10: Maintenance of the ship and equipment

Software updates, security patches, configuration management, backup verification, account administration, malware protection, network-device maintenance and vendor-access controls should be treated as managed maintenance activities.

Where patching is delayed because of operational or vendor constraints, the resulting risk and any compensating controls should be documented.

ISM 11: Documentation

Cybersecurity policies, risk assessments, inventories, diagrams, incident-response plans and related records should be controlled as SMS documentation where applicable.

Access should be restricted according to need, changes should be authorized, obsolete documents should be removed and sensitive technical information should be protected.

ISM 12: Company verification, review and evaluation

Internal audits should examine cyber risk management, and management reviews should consider its effectiveness.

The audit should not be reduced to confirming that a cybersecurity policy exists. It should test whether procedures are being followed, records are reliable, responsibilities are understood and controls remain appropriate for the actual vessel and shore environment.

If cyber risk appears in the SMS but touches none of these areas, it has not been meaningfully integrated. It has been appended.

Shipowner and operator responsibilities

The current Guidelines state that effective cyber risk management should begin at senior-management level and be embedded across the organisation.

For ISM purposes, formal responsibility rests with the Company. Meaning the owner or another organisation, such as a ship manager, that has assumed responsibility for operating the ship and agreed to discharge the duties imposed by the ISM Code.

In practical terms, the Company should:

  1. Establish accountability. A person or entity should be designated to oversee the planning, resourcing and execution of cybersecurity activities, with sufficient authority, support and competence.
  2. Maintain an accurate asset inventory. The company should identify onboard digital systems, software, hardware, network connections, critical data, internal and external dependencies and connections to shore or third-party infrastructure. An incomplete inventory undermines every subsequent control because a company cannot reliably protect, monitor, maintain or recover systems it has not identified. In my professional observation, asset visibility remains one of the most significant implementation weaknesses across maritime environments.
  3. Conduct and maintain risk assessments. Risk assessment should cover systems, services, assets, data and capabilities whose sudden failure or compromise could affect safe operations, human safety, the vessel or the environment. It should also cover software and hardware supply-chain exposure.
  4. Implement proportionate controls. The controls described in the IMO Guidelines should be considered a baseline, supplemented where the assessed risk justifies stronger measures.
  5. Train shore and vessel personnel. Training should be appropriate to role and should be supported by familiarisation, exercises and periodic testing of knowledge.
  6. Prepare for incidents. The company should be able to detect, respond to, record, report and recover from cyber incidents while maintaining or restoring safe operations.
  7. Analyze and improve. Incidents, near-misses, audit findings and identified weaknesses should lead to root-cause analysis, corrective action and fleet-wide learning where relevant.
  8. Protect cyber documentation. Risk assessments, network diagrams, asset inventories, credentials, incident records and response plans may themselves be sensitive. They should be protected against unauthorised disclosure, modification or destruction.

Where an owner transfers ISM responsibility to a third-party ship manager, the manager may become the formal ISM Company. That does not mean the owner has no remaining exposure.

The owner continues to face asset, commercial, contractual, financing, insurance and reputational consequences if the manager’s cyber risk-management arrangements are ineffective. Owners should therefore exercise appropriate due diligence rather than assume that the existence of a manager’s DOC automatically proves effective cyber implementation across the managed fleet.

Evidence of effective implementation

Article content
A policy shows intent; evidence proves implementation.

This is where the framework becomes operational. The ISM framework distinguishes between observations, non-conformities and major non-conformities.

A major non-conformity includes an identifiable deviation posing a serious threat to personnel, ship safety or the environment and requiring immediate corrective action. The definition also covers a lack of effective and systematic implementation of an ISM Code requirement.

This does not mean that every cyber weakness automatically constitutes a major non-conformity.

The classification of a finding depends on objective evidence, the applicable requirement, the seriousness of the condition, the consequences, the extent of the failure and the Administration’s or recognized organization’s audit procedures.

However, where cyber risk management has been included within the approved SMS but is not being effectively or systematically implemented, the issue is capable of supporting an ISM finding. In serious cases, and depending on the circumstances, escalation to a major non-conformity is possible.

The Port State Control perspective

Article content
Port State Control tests implementation—not policy wording.

The US Coast Guard’s published work instruction for Marine Inspectors and Port State Control Officers, CVC-WI-027 (document), provides a useful example of how this logic may be applied. The current revision, CVC-WI-027(3), dates from October 2023.

The guidance directs inspectors to consider cyber hygiene and, where clear grounds exist, conduct a more detailed examination.

It also makes two important points.

  1. Poor cyber hygiene or a more detailed inspection does not automatically mean that an ISM deficiency exists.
  2. Where objective evidence demonstrates that the vessel has failed to implement the cyber risk-management provisions of its own SMS, a deficiency may be issued against the relevant SMS requirement. Inspectors are not instructed to impose their preferred cybersecurity checklist on the vessel; the test is whether the company’s approved management system is being effectively implemented.

For foreign vessels, the work instruction distinguishes between several possible situations. Where an expanded examination reveals that cyber risk management was not incorporated into the vessel’s SMS by the required DOC verification, the PSCO may issue an action code 17 deficiency requiring rectification and an external audit before departure. Where cyber risk management has been incorporated but the vessel has failed to implement it, the PSCO may issue the related operational deficiency together with an ISM deficiency carrying action code 40 [rectification before the vessel’s next U.S. port after sailing foreign] or action code 50, requiring rectification within 30 days and an internal audit focused on the vessel’s cyber risk management. Deficiency Codes List.

Where a serious failure to implement cyber risk management directly resulted in a cybersecurity incident affecting ship operations, diminished safety or security, or increased environmental risk, the work instruction contemplates action code 30 [ship detained] following concurrence from the Officer in Charge, Marine Inspection, together with an external audit before release from detention.

Publicly verifiable cases of Port State Control detention arising specifically and exclusively from cyber-related ISM findings remain difficult to identify in the open record.

-= No such claim is made here =-.

The available enforcement and audit consequences are nevertheless documented and real. Depending on the circumstances, they may include an operational deficiency, an ISM deficiency, a non-conformity or major non-conformity, additional internal or external audit requirements, certificate-related action or detention.

What objective evidence looks like

Auditors and inspectors do not verify intent. They verify implementation through evidence.

Relevant evidence may include:

  • An accurate onboard asset inventory reconciled against what is actually installed.
  • Current network diagrams and records of external connections.
  • Cyber risk assessments that are dated, approved, reviewed and updated following material changes.
  • Records demonstrating control of privileged and inactive accounts.
  • Training and familiarization records, including OT-specific training where applicable.
  • Exercise and drill records showing that lessons were identified and addressed.
  • Cyber incident and near-miss records.
  • Root-cause analyses and corrective-action close-outs.
  • Internal audit reports that meaningfully assess cyber risk.
  • Management-review records demonstrating senior-level oversight.
  • Vendor remote-access approvals, access logs and termination records.
  • Software, configuration, patching and backup records.
  • Evidence that restoration procedures and backups have been tested.
  • Records demonstrating that removable media and portable devices are controlled.
  • Evidence of hardware, software and supplier approval processes.

A policy document with no supporting records is not evidence of implementation. It is evidence of intent.

How the IMO framework fits with the ISM Code, IACS E26/E27 and industry guidance

Article content
One framework, multiple layers.

The maritime cyber framework is best understood as several interacting layers, each performing a different function.

Confusing those layers is the source of many compliance errors.

1. The ISM Code: the mandatory management framework

The ISM Code is mandatory for applicable companies and ships through SOLAS chapter IX.

It requires a structured Safety Management System, assessment of identified risks, appropriate safeguards, defined responsibilities, emergency preparedness, maintenance, reporting, documentation, internal verification and management review.

MSC.428(98) confirms that cyber risk management belongs within that framework. The ISM obligation applies according to the scope of SOLAS chapter IX and the relevant flag-State implementation. It is not limited to vessels constructed after a particular cyber-regulation date.

2. MSC-FAL.1/Circ.3/Rev.4: the current IMO benchmark

The current Guidelines are formally non-mandatory and provide high-level recommendations.

They nevertheless offer the most current IMO benchmark for understanding how maritime cyber risks should be governed, identified, protected against, detected, responded to and recovered from.

Their use is not a substitute for considering flag-State requirements, national law, class rules or the company’s own risk profile.

The Guidelines themselves direct users toward relevant Administration requirements, international standards and industry best practices for more detailed implementation.

3. IACS UR E26 and E27: class requirements for covered new construction

IACS UR E26 addresses the cyber resilience of the ship as a collective and integrated entity.

IACS UR E27 addresses the cyber resilience of onboard systems and equipment, including requirements intended to ensure that security capabilities are incorporated at the design and manufacturing stages.

The Rev.1 versions are to be uniformly implemented by IACS member societies for covered ships contracted for construction on or after “1 July 2024″. They may be used for other ships as non-mandatory guidance.

Their mandatory vessel scope includes:

  • Passenger ships, including passenger high-speed craft, engaged in international voyages.
  • Cargo ships of 500 GT and above engaged in international voyages.
  • High-speed craft of 500 GT and above engaged in international voyages.
  • Mobile offshore drilling units of 500 GT and above.
  • Specified self-propelled mobile offshore units engaged in construction activities, such as wind-turbine installation, maintenance and repair units, crane units, drilling tenders and accommodation units.

The boundary between the IMO and IACS regimes is important.

  • E26 and E27 do not retrospectively impose their mandatory new-construction requirements across the existing fleet.
  • A vessel built in 2015 will ordinarily fall outside their mandatory application. However, the requirements or equivalent controls may still be used voluntarily, through class notations, retrofits, owner specifications, charterer requirements, contractual arrangements or company standards.
  • That vessel may also remain subject to cyber risk-management requirements through the ISM Code, flag-State rules, port-State requirements and other applicable legislation.
  • Equally, a newbuilding constructed to meet E26 and E27 is not automatically compliant with every cyber-related ISM obligation.
  • Class verification of design, integration, onboard systems and technical resilience is distinct from the company’s responsibility to integrate cyber risk into its SMS and demonstrate effective implementation during operations.

At the same time, it would be inaccurate to describe E26 as relevant only at delivery.

E26 contains operational-life requirements for shipowners, including management of technical and organizational security measures, maintenance of documentation, operational procedures, periodic training and drills, updating of cyber-resilience test procedures and verification through annual and other class surveys.

The two regimes therefore overlap operationally, but they remain different in legal basis, scope, verification mechanism and purpose.

4. Industry guidance: where much of the practical detail lives

The Guidelines on Cyber Security Onboard Ships, Version 5, published on “14 November 2024” and supported by major shipping organizations, remains an important industry implementation reference.

The current IMO Guidelines also refer users to:

  • ISO/IEC 27001.
  • IACS UR E26 and UR E27.
  • Consolidated IACS Recommendation No. 166 on Cyber Resilience.
  • The NIST Cybersecurity Framework 2.0.
  • The IAPH Cybersecurity Guidelines for Ports and Port Facilities.
  • The IAPH Cyber Resilience Guidelines for Emerging Technologies in the Maritime Supply Chain.

These materials are not all IMO instruments, and they do not all carry the same legal status.

Their value is that they provide technical, organizational and procedural detail that a short, high-level IMO circular cannot reasonably contain.

Additional national and regional requirements

Other legal regimes may apply alongside the IMO framework.

The US Coast Guard’s final rule on cybersecurity in the Marine Transportation System was published on “17 January 2025” and became effective on “16 July 2025”. It establishes specific requirements for covered U.S.-flagged vessels, regulated facilities and Outer Continental Shelf facilities.

It is separate from the USCG Port State Control work instruction used to assess foreign vessels under the ISM framework.

Within the European Union, NIS2 may apply to qualifying entities in the water-transport sector, depending on their size, activities, jurisdiction and national transposition.

NIS2 obligations do not replace the ISM Code. They address the cybersecurity risk-management and reporting responsibilities of in-scope entities under a separate legal framework.

The applicable maritime cyber obligations therefore depend on a combination of factors, including vessel type, tonnage, voyage profile, contract or construction date, flag, class, operating jurisdiction, company status and connections to regulated infrastructure.

They do not depend on vessel type alone, and they cannot be demonstrated through a single certificate.

Where this is heading

Article content
The next milestone is the Maritime Cyber Code.

The current IMO architecture consists of a mandatory management Code, an interpretive resolution and non-mandatory supporting Guidelines. That framework has raised maritime cyber risk to the level of formal safety management, but it remains structurally fragmented.

The IMO has now begun developing a dedicated, goal-based and non-mandatory Maritime Cyber Code.

At MSC 110 in June 2025, the Maritime Safety Committee agreed that a non-mandatory cybersecurity code should be developed using a risk-management-oriented approach. FAL 50 subsequently recorded that the Working Group established during MSC 110 had agreed that the ISPS Code was not an appropriate approach for developing comprehensive cybersecurity arrangements within IMO, and that a goal-based, non-mandatory Maritime Cyber Code offered a more comprehensive pathway.

At FAL 50 in March 2026, the Facilitation Committee approved a new output for the development of a non-mandatory Maritime Cyber Code, with a target completion year of 2028. It invited the Maritime Safety Committee to participate as an associated organ, approved the road map for the approval of the Code, approved the formation of an Intersessional Correspondence Group to develop it, and approved an intersessional joint FAL-MSC expert group on the Code, to meet in 2027.

MSC 111, in May 2026, agreed to participate as an associated organ and concurrently approved the road map.

Both the Correspondence Group and the joint FAL-MSC expert group remain subject to consideration and endorsement by the IMO Council at its 137th session. The Correspondence Group is to report to FAL 51, provisionally scheduled for February 2027, where a dedicated working group on the Code is anticipated.

The Code is intended to be goal-based and to address the wider maritime digital ecosystem, including computer-based systems supporting the safe operation of port facilities, shipping and the ship-to-shore interface.

During FAL 50, one of the views recorded in support of the proposal was that the Code’s initially non-binding nature, potentially coupled with an experience-building phase, would allow States and industry to build capability, test solutions and identify gaps without immediately creating a fully prescriptive global regime.

Whether the Maritime Cyber Code will eventually form the basis of a mandatory IMO instrument has not been decided.

The development of the Maritime Autonomous Surface Ships Code demonstrates one possible pathway. Beginning with a non-mandatory instrument, followed by experience building and later consideration of mandatory status.

Applying the same trajectory to the Maritime Cyber Code, however, would currently be an inference rather than an agreed IMO position. The only firm conclusion is that the regulatory direction is toward greater structure, clearer governance and broader harmonization.

The intervening period should not be treated as a grace period.


Article content
Cyber resilience is demonstrated through implementation—not documentation.

MSC.428(98) is a short resolution with a precise purpose.

It confirms that maritime cyber risk belongs within the ISM risk-management framework and should be addressed through the approved Safety Management System.

It did not establish a separate cyber certificate, and the obligation cannot be satisfied merely by stating that the company “complies with MSC.428.”

The operational instrument to read today is MSC-FAL.1/Circ.3/Rev.4.

It carries forward a significantly stronger framework than the one against which many early maritime cybersecurity policies were developed:

  • A formal governance function.
  • A designated accountable person or entity.
  • Controls described as minimums.
  • Identification of onboard systems and dependencies.
  • IT and OT network segmentation.
  • Supply-chain risk management.
  • Annual basic cybersecurity training.
  • OT-specific competence.
  • Periodic testing through drills and exercises.
  • Incident recording and reporting.
  • Root-cause analysis.
  • Periodic assessment of effectiveness.
  • Protection of cyber risk-management documentation.

The bar in 2026 is not the existence of documentation. It is demonstrable implementation.

The ISM framework places particular weight on objective evidence and on the effective and systematic operation of the company’s management system. Companies most exposed are therefore not necessarily those without a cybersecurity policy.

They are the companies with a detailed policy, a valid DOC and insufficient evidence that the policy is understood, resourced, followed, tested and improved across their shore organisation and fleet.

The IMO’s development of a non-mandatory Maritime Cyber Code will bring another stage of maturity to this framework.

Companies that have genuinely integrated cyber risk into their Safety Management Systems will be able to absorb that development as an evolution. Companies that have built only a compliance artefact may eventually find themselves rebuilding their arrangements under audit, regulatory, contractual or incident-driven pressure. On someone else’s timeline.


Sources

This article draws primarily on the following official instruments, regulatory documents and recognized industry publications:

  • IMO Resolution MSC.428(98), Maritime Cyber Risk Management in Safety Management Systems, adopted 16 June 2017.
  • MSC-FAL.1/Circ.3/Rev.4, Guidelines on Maritime Cyber Risk Management, issued 28 May 2026.
  • MSC-FAL.1/Circ.3/Rev.3, Guidelines on Maritime Cyber Risk Management, issued 4 April 2025.
  • The International Safety Management Code and SOLAS chapter IX.
  • IMO Resolution A.1188(33), 2023 Guidelines on Implementation of the International Safety Management (ISM) Code by Administrations, adopted 6 December 2023.
  • IACS Unified Requirement E26 Rev.1, Cyber Resilience of Ships.
  • IACS Unified Requirement E27 Rev.1, Cyber Resilience of On-Board Systems and Equipment.
  • Consolidated IACS Recommendation No. 166 on Cyber Resilience.
  • IACS Recommendation No. 41 Rev.5, Guidance for Auditors to the ISM Code, October 2019.
  • Report of the Facilitation Committee on its fiftieth session (FAL 50/20), including the road map for the approval of the Maritime Cyber Code.
  • IMO meeting outcomes for MSC 108, MSC 110, FAL 49 and MSC 111.
  • US Coast Guard Work Instruction CVC-WI-027(3), Vessel Cyber Risk Management Work Instruction, 11 October 2023.
  • The US Coast Guard final rule, Cybersecurity in the Marine Transportation System, published 17 January 2025 and effective 16 July 2025.
  • The Guidelines on Cyber Security Onboard Ships, Version 5, published 14 November 2024.
  • Directive (EU) 2022/2555 (NIS2).
  • IAPH cybersecurity and cyber-resilience guidance.
  • IMO Resolution A.1188(33), 2023 Guidelines on Implementation of the International Safety Management (ISM) Code by Administrations, adopted 6 December 2023.
  • IACS Recommendation No. 41 Rev.5, Guidance for Auditors to the ISM Code, October 2019.
  • US Coast Guard Work Instruction CVC-WI-027(3), Vessel Cyber Risk Management Work Instruction, 11 October 2023.

No commercial relationship exists with any vendor, classification society or organisation referenced in this article.

Publicly documented Port State Control detentions arising specifically and exclusively from cyber-related ISM findings could not be reliably verified in the open record at the time of writing, and no such claim is made.


Maritime Cyber Insights is an independent series of articles and posts examining the regulatory, operational, and strategic dimensions of cyber risk across the global maritime industry. It is published by Alexandros Engelen, a Cybersecurity Strategist specializing in maritime cyber risk, with a focus on helping shipping companies, maritime stakeholders, and decision-makers understand how cyber obligations translate into practical risk management.

Maritime Cyber Compliance Is Not a Vessel-Type Exercise

A recurring question surfaces in fleet reviews and board-level risk discussions: Which cyber requirements apply to my ships?

It is almost always asked in terms of vessel type. As though a tanker, a bulk carrier and a passenger ferry each carry a distinct cyber rulebook. They do not. The instinct is understandable, because so much of maritime regulation is stratified by ship type, but for cyber risk it produces a misread of where the obligations actually sit. Owners who scope their exposure by hull category tend to over-engineer for some vessels and leave material gaps on others.

The reality is that maritime cyber obligations attach to a different set of variables:

  • the flag a ship flies
  • its gross tonnage and voyage type
  • its build date
  • the jurisdictional footprint of the operating company
  • [for a growing share of the commercial fleet] the vetting regimes a vessel must satisfy to earn employment.

Vessel type modifies this picture at the edges, sometimes sharply. But it is not the organising principle. Understanding that distinction is the difference between a defensible compliance posture and a matrix full of false confidence.

This article maps the obligations that reach the whole commercial fleet, then isolates the points where vessel type genuinely changes what an owner must do.

The horizontal layer

The following instruments land on essentially every commercial vessel a shipowner operates, keyed not to type but to flag, tonnage and trade.

Article content
Maritime cyber compliance is built across the fleet, not around vessel type.
  1. IMO Resolution MSC.428(98) and the ISM Code. This is the anchor. Adopted in June 2017, the resolution directs administrations to ensure that cyber risks are addressed within the Safety Management System, as defined by the ISM Code, no later than the first annual Document of Compliance verification after 1 January 2021. The resolution is not, in itself, a mandatory instrument. Its force comes from the linkage to the ISM Code under SOLAS Chapter IX: cyber risk management becomes an effective condition of the company’s Document of Compliance and each ship’s Safety Management Certificate. This is the obligation that reaches every ISM-regulated vessel in a fleet, and it is verified, not merely encouraged.
  2. IMO Guidelines MSC-FAL.1/Circ.3/Rev.3. The current high-level guidance. Revision 3 was approved by the Maritime Safety Committee at its 108th session in May 2024 and issued in April 2025. It is materially more current than its predecessor: it now references IACS Unified Requirements E26 and E27 as recommended standards and aligns with version 2.0 of the National Institute of Standards and Technology (NIST) Cybersecurity Framework. The guidance remains non-mandatory, but it defines what “adequate” cyber risk management looks like when a flag or port State assesses SMS compliance. Treating it as optional is a misjudgement of how it is used in practice.
  3. IACS Unified Requirements E26 and E27. These apply to ships contracted for construction on or after 1 July 2024. UR E26, “Cyber Resilience of Ships,” addresses the vessel as an integrated system; UR E27, “Cyber Resilience of On-Board Systems and Equipment,” addresses individual computer-based systems and pushes obligations onto equipment suppliers. The original versions, carrying a 1 January 2024 application date, were withdrawn and superseded. One feature of the revised requirements matters directly to any type-based analysis: applicability is categorised as mandatory or non-mandatory depending on vessel type and size, a deliberate accommodation for smaller and non-conventional vessels. This is the one place in the horizontal layer where type is written into the instrument itself.
  4. EU Cyber Resilience Act. A further supplier-side development to watch is the EU Cyber Resilience Act. It does not regulate a ship as such, but it will increasingly affect the manufacturers and suppliers of connected software, hardware and digital products placed on the EU market. For maritime owners, its practical importance will sit in procurement, lifecycle support, vulnerability handling and supplier assurance.
  5. Classification society cyber rules. Each IACS member operationalizes E26 and E27 through its own class notations: DNV’s Cyber Secure, ABS’s CyberSafety, and the equivalent frameworks from Bureau Veritas, Lloyd’s Register and ClassNK. For an owner, these function as class conditions attached to newbuildings and, increasingly, to retrofits and voluntary notations on existing tonnage.
  6. Flag State requirements. Every flag transposes the International Maritime Organization framework into national law, and some legislate beyond it. The obligation an owner actually carries therefore varies by flag, and a mixed-flag fleet carries a correspondingly mixed obligation set. This is best treated as a per-flag variable rather than a constant.
  7. The NIS2 Directive (EU 2022/2555). This attaches to the relevant organisation and service, not to the individual hull. In maritime, NIS2 sits within the transport sector and covers inland, sea and coastal passenger and freight water transport companies, port managing bodies including port facilities and entities operating works and equipment within ports, and operators of Vessel Traffic Services. Crucially, the Directive expressly excludes the individual vessels operated by those water transport companies. This is a key distinction for shipping groups: NIS2 captures the relevant legal entity and service, not the ship itself. The directive applies broadly to medium-sized and larger entities under the EU SME Recommendation, subject to national implementation and specific NIS2 exceptions; smaller operators may still be captured where, for example, they are the sole provider in a Member State of a service essential to critical societal or economic activity, or where disruption could create significant public-safety, public-security, public-health, systemic or cross-border impact. Obligations include structured cyber-risk management, supply-chain security, incident handling, business continuity, backup and disaster recovery, vulnerability handling, access control, asset management, cyber hygiene, management-body oversight and training, and a staged incident-reporting regime: an early warning within 24 hours, an incident notification within 72 hours, and a final report generally within one month. Transposition was due on 17 October 2024. As of July 2026, the European Commission has referred Ireland, Spain, France and the Netherlands to the Court of Justice of the European Union for failing to notify full transposition, so the practical obligation must be checked against the national implementing law in each Member State where the relevant maritime entity is established or provides in-scope services.
  8. The USCG rule on Cybersecurity in the Marine Transportation System. Codified at 33 CFR 101.600 et seq. and effective 16 July 2025, this applies to U.S. flagged vessels, Outer Continental Shelf facilities, and facilities regulated under the Maritime Transportation Security Act. Implementation is phased: reportable cyber incidents must go to the National Response Center immediately from the effective date; personnel training under 33 CFR 101.650 is required by 12 January 2026 and annually thereafter; and Cybersecurity Officer designation, a Cybersecurity Assessment, and submission of a Cybersecurity Plan are required by 16 July 2027. Foreign-flag owners should not read this as direct application of the USCG cyber rule. Subpart F expressly excludes foreign-flagged vessels subject to 33 CFR Part 104. The practical exposure is different: foreign-flagged vessels calling at U.S. ports may still face Port State Control scrutiny where weak cyber-risk management indicates poor implementation of the ISM Code or a deficient Safety Management System. In other words, the rule does not directly regulate every foreign-flagged vessel trading to the United States, but it raises the inspection and deficiency risk for owners whose cyber controls cannot withstand ISM-based scrutiny.
  9. Industry guidance and commercial obligations. Not regulation, but enforced through the market with comparable effect. The BIMCO and International Chamber of Shipping Guidelines on Cyber Security Onboard Ships [version 5, published at the end of 2024] is the “de facto” SMS reference. On the commercial side, charterer and vetting regimes now embed cyber requirements that are, for many owners, more operationally binding than the statutory floor: OCIMF’s SIRE 2.0 for tankers and RightShip for dry bulk both make cyber posture a condition of employment.

Where vessel type genuinely changes the picture

Against that horizontal baseline, type-specific factors alter the obligation set in identifiable ways. The following matrix isolates them.

Article content
Vessel type modifies cyber obligations. It does not define them.

The pattern is consistent. Passenger ships attract heightened obligation through life-safety exposure and universal ISM applicability. Tankers and bulk carriers pick up their most binding cyber requirements not from statute but from vetting. Offshore units are named directly in U.S. regulation. And the smallest vessels [the segment most likely to be assumed out of scope entirely] are handled explicitly by the E26 and E27 scoping and may still be reached through the operating company under NIS2. Type matters. It simply does not organise the framework.

Scoping a fleet correctly

Article content
Correct scoping turns maritime cyber compliance from assumption into evidence.

In complex shipping structures, the first scoping exercise should identify the relevant legal entity: registered owner, Document of Compliance holder, technical manager, operator, bareboat charterer, port operator, Vessel Traffic Services operator or group parent. Cyber obligations often attach to the entity performing the regulated service, not simply to the asset-owning company.

For any individual hull, the applicable obligation set is a product of five variables. The flag determines which IMO transposition and any national additions apply. Gross tonnage and voyage type determine whether the ISM and ISPS regimes engage at all. The shipbuilding contract date determines whether IACS UR E26 and E27 apply, since the revised requirements apply to ships contracted for construction on or after 1 July 2024. The company’s EU establishment, size, role and criticality determine whether NIS2 applies and whether the entity is treated as essential or important under national implementing law. And the U.S. nexus [whether by U.S. flag, OCS operation, MTSA-regulated facility, or foreign-flag calls at U.S. ports exposed to Port State Control scrutiny] determines the direct or practical reach of the Coast Guard framework.

A defensible maritime cyber compliance file should not only list applicable regulations. It should contain evidence: the SMS cyber-risk procedure, vessel IT/OT asset inventory, remote-access register, backup and restore records, crew familiarisation records, incident escalation matrix, supplier access controls, class/newbuilding cyber documentation, and vetting responses for SIRE 2.0 or RightShip where applicable. In practice, this is what turns cyber compliance from a policy statement into something that can survive audit, inspection or charterer review.

Run those five variables across a fleet and the picture collapses into something far more precise than a hull-by-hull rulebook. Two sister ships under different flags can carry different obligations; a newbuilding and an existing vessel of identical type will diverge on E26 and E27; and a modest coastal operator with a critical port role can fall within NIS2 while a larger deep-sea owner structured differently may face a different classification or obligation set. This is why type-based scoping fails: it holds the wrong variable constant.

The trajectory

The direction of travel is toward convergence. MSC-FAL.1/Circ.3/Rev.3 now points explicitly at the IACS Unified Requirements and at NIST CSF 2.0, drawing the IMO guidance, class rules and a recognised control framework into closer alignment. NIS2 and the USCG rule both introduce something the IMO framework has historically lacked: enforceable incident reporting on defined timelines, and [in the case of NIS2] personal accountability at senior-management level. The regulatory floor is rising, and it is rising fastest where operational technology, supply-chain dependency and shore integration are densest. Those are themes that recur across the threat landscape rather than attaching to any single class of vessel, which is precisely why the obligation framework is moving away from type-based logic and toward a posture assessed at the level of systems, company and trade.

For owners, the practical implication is straightforward. Scope by the variables that actually govern [flag, tonnage, trade, build date and jurisdictional nexus] and treat vessel type as a modifier rather than the map. A compliance matrix built the other way round will look complete and still leave the fleet exposed.


Maritime Cyber Insights is an independent series of articles and posts examining the regulatory, operational, and strategic dimensions of cyber risk across the global maritime industry. It is published by Alexandros Engelen , a Cybersecurity Strategist specializing in maritime cyber risk, with a focus on helping shipping companies, maritime stakeholders, and decision-makers understand how cyber obligations translate into practical risk management.


Sourcing: This analysis draws on primary and authoritative sources including IMO Resolution MSC.428(98) and Guidelines MSC-FAL.1/Circ.3/Rev.3; IACS Unified Requirements E26 and E27 and associated IACS press material; Directive (EU) 2022/2555 (NIS2) and European Commission and ENISA guidance; the U.S. Coast Guard final rule on Cybersecurity in the Marine Transportation System (33 CFR 101.600 et seq.; 90 Fed. Reg. 6298); the ISM and ISPS Codes under SOLAS; and the BIMCO/ICS Guidelines on Cyber Security Onboard Ships (v5). Regulatory positions current as of publication; NIS2 national transposition and the USCG U.S.-flag implementation timeline remain in movement and should be verified against the primary texts for any specific jurisdiction.

What are classification societies in the maritime industry ?

If you work in shipping, you encounter their names constantly:
– DNV
– Lloyd’s Register
– ABS Group
– Bureau Veritas
– ClassNK – Nippon Kaiji Kyokai
– RINA

But outside technical circles, their actual role is often misunderstood.

Classification societies are independent technical organizations that establish and apply rules for the design, construction, and operation of ships and offshore structures.

When a vessel is built and maintained in accordance with a society’s rules, it is said to be “in class.”

This is not just a formality.

Class status supports a vessel’s technical credibility, insurability, chartering acceptance, and the confidence that owners, operators, charterers, insurers, financiers, port states, and flag states place in its condition.

Their role spans the full lifecycle of a vessel:
– Design approval before construction begins.
– Surveys during construction.
– Periodic surveys during operation.

Verification that the vessel continues to comply with applicable class requirements throughout its service life.

It is also important to distinguish between classification and statutory certification.

Classification societies apply their own technical rules. However, when authorized by a flag State, they may also act as Recognized Organizations, carrying out statutory surveys and certification on behalf of that flag administration.

This is where their influence becomes even more important.

The twelve IACS member societies coordinate through the International Association of Classification Societies, which develops unified requirements and interpretations adopted across member fleets.

And this is where the cyber dimension becomes unavoidable.

IACS Unified Requirements E26 and E27, applicable to new ships contracted for construction on or after 1 July 2024, extend classification thinking into cyber resilience.

1. UR E26 focuses on the cyber resilience of the ship as a whole.
2. UR E27 focuses on the cyber resilience of onboard systems and equipment.

In other words, the same institutional framework that has governed hull, machinery, and safety-related technical assurance for more than a century is now shaping the baseline for cyber integrity at sea.

Understanding what classification societies do is the first step to understanding who sets many of the technical rules maritime cyber security must now answer to.

What are IACS UR E26 and E27 ?

IACS Unified Requirements E26 and E27 appear repeatedly in maritime cyber security discussions. Often referenced, less often explained.

They mark the point where IACS, through its member classification societies, extended the logic of technical assurance into cyber resilience.

  1. So what do these two requirements actually do?
  2. And why does the distinction between them matter?

They address the problem from two different altitudes.

UR E26: Cyber Resilience of Ships

E26 treats the vessel as a single integrated system.
It establishes a minimum set of requirements for the cyber resilience of a ship across its design, construction, commissioning, and operational life.

Its structure follows five key functions:

  • Identify
  • Protect
  • Detect
  • Respond
  • Recover

Its concern is the whole vessel. How onboard IT and OT systems are integrated, documented, tested, surveyed, and maintained as part of a cyber-resilient ship.

The key audience includes

  1. Shipyards
  2. System integrators
  3. Owners
  4. Operators
  5. Classification societies

UR E27: Cyber Resilience of On-Board Systems and Equipment

E27 works one level down.
It defines the minimum security capabilities that onboard systems and equipment must have in order to be considered cyber resilient.

This requirement is aimed primarily at equipment manufacturers and third-party suppliers.

It connects maritime cyber resilience with established industrial control system security principles, including references to IEC 62443-3-3 and IEC 62443-4-1.

In simple terms:

  • E27 hardens the components.
  • E26 governs the ship they are assembled into

Together, they create a layered model for maritime cyber resilience.

One point is widely misunderstood. The versions in force today are not the original editions.

IACS withdrew the first editions before they entered into force and replaced them with revised requirements:

  • UR E26 Rev.1
  • UR E27 Rev.1

These apply to ships contracted for construction on or after 1 July 2024.

Applicability is also now split between mandatory requirements and non-mandatory guidance depending on vessel type and size.

For anyone benchmarking against E26 or E27, working from the withdrawn 1 January 2024 editions is not a small administrative mistake.

It is a live compliance risk.

The keys to the edge: FortiBleed and the silent exposure of maritime network perimeters

Maritime Cyber Threat Briefing #12

Most maritime cyber incidents announce themselves. A ransomware note appears, an ECDIS position drifts, a booking portal goes dark. FortiBleed is different. It is a compromise that leaves no obvious mark on the victim, that touches devices which appear fully patched and operational, and that has already placed working administrative keys to hundreds of shipping and offshore networks into a criminal marketplace. For the maritime sector, it is one of the most consequential exposures of the year precisely because so many of the affected operators still do not know they are affected.

This briefing sets out what FortiBleed is, why it defeated organizations that believed they had done the right thing by patching, what maritime-specific research has uncovered about the industry’s exposure, and the concrete steps owners, managers, ports and vendors should take now. It builds directly on ground covered in Briefing #3 (SSL-VPN exploitation), Briefing #4 (VSAT and satcom vulnerabilities) and Briefing #11 (Volt Typhoon and the strategic pre-positioning of maritime infrastructure), because FortiBleed sits at the intersection of all three.

Background: the firewall is the front door

Article content
The Perimeter Breach Becomes a Fleet-Wide Risk

Fortinet‘s FortiGate firewalls and SSL-VPN gateways are among the most widely deployed perimeter security devices in the world, and the maritime sector is no exception. They guard the boundary between the public internet and the internal networks of shipping companies, ship managers, offshore contractors, yards and ports. On many vessels, a FortiGate also sits between the satellite link and the onboard network, making it the single most important gatekeeper between a ship and the shore.

That placement is exactly what makes these devices a strategic target. An attacker who holds valid administrator credentials for a perimeter firewall does not need an exotic exploit. They can modify firewall rules, read and redirect VPN traffic, create hidden accounts, disable logging, and use the device as a quiet foothold from which to move deeper into the network.

The firewall is the front door, and FortiBleed is, in effect, a set of copied keys.

The campaign takes its name from Heartbleed, but the comparison is only rhetorical. This is not a single dramatic software vulnerability. It is a large-scale credential theft and validation operation, first surfaced in mid-June 2026 by security researcher Volodymyr “Bob” Diachenko, who discovered an exposed attacker-controlled server, and independently verified by Kevin Beaumont and Hudson Rock. Multiple threat-intelligence firms, including SOCRadar® Extended Threat Intelligence, Arctic Wolf, Recorded Future‘s “Insikt Group” and Bitsight, have since analysed the dataset and the attacker toolchain.

The mechanism: why patched devices stayed exposed

Article content
Patching the Device Did Not Neutralize the Exposure

The technical heart of FortiBleed is a credential-storage weakness that survived a security upgrade. It is worth understanding clearly, because the lesson generalises well beyond Fortinet.

For years, FortiGate devices stored administrator passwords as salted SHA-256 hashes. SHA-256 is fast to compute, which is a virtue for many uses and a liability for password storage. A modern graphics-processing cluster can test billions of guesses per second against it, so even long, complex passwords fall in hours. Fortinet addressed this by migrating to PBKDF2, a deliberately slow, cracking-resistant algorithm, across FortiOS versions 7.6.1, 7.2.11 and 7.4.8 released between late 2024 and mid-2025.

The migration carried a critical condition. When a device was upgraded, existing administrator passwords were not automatically re-hashed. They remained in the old, weak SHA-256 format until each administrator logged in again after the upgrade, which triggered the rehash. Worse, the older hash was retained in a hidden configuration field for backward compatibility even after a password was updated. The consequence is the defining feature of this incident: an organisation could apply the patch, believe the matter closed, and still be carrying crackable legacy hashes inside its live configuration.

According to Fortinet’s own analysis, this is not a new vulnerability but the reuse of credentials from prior incidents combined with aggressive brute-forcing against devices with weak password hygiene and no multi-factor authentication. That framing matters for the maritime reader in one specific way: because it is not classed as a formal vulnerability, there is no new CVE and no patch that “fixes” FortiBleed. The remediation is operational discipline, not a software update. A point returned to below.

The attack chain, and why it reads as an industrial operation

The value of FortiBleed to an attacker comes from the systematic way the operation converts exposure into verified, sellable access.

The publicly documented chain, reconstructed from the attackers’ own exposed tooling, runs roughly as follows.

Configuration files were exfiltrated from internet-facing FortiGate devices. The precise initial-access method is unconfirmed, but analysts have identified CVE-2026-24858, a critical FortiCloud single-sign-on authentication bypass added to Cybersecurity and Infrastructure Security Agency‘s Known Exploited Vulnerabilities catalogue earlier in 2026, as the leading candidate. Those configuration files contained the legacy password hashes, and in some cases plaintext credentials outright.

The hashes were then cracked offline on a dedicated 45-GPU cluster managed through Hashtopolis. Because the legacy SHA-256 storage is computationally cheap, this stage is fast and effective. In parallel, the operators replayed enormous volumes of previously leaked and infostealer-harvested credentials against exposed devices. Threat intelligence points to roughly 1.16 billion login attempts against more than 320,000 FortiGate targets, alongside billions more directed at database servers. The credential pools were curated specifically from earlier Fortinet leaks, including the near-500,000 VPN accounts exposed in 2021 and the Belsen Group configuration dump of 2022, meaning the attackers were testing passwords that had already proven valid on Fortinet devices.

Every successful login was verified against the live device and then profiled. This is the signature of an initial access broker: compromised access was catalogued by country, industry, company revenue and headcount, packaged for resale to ransomware syndicates and other buyers. On devices they controlled, the operators deployed an on-box packet-sniffing tool that captured SSL-VPN and other credentials directly from passing traffic as legitimate users signed in, feeding fresh logins back into the operation. From the firewall, the documented path continues into Active Directory and, where segmentation is weak, onward toward operational systems.

Article content
This Is Not Opportunistic. This Is Industrial.

I have deliberately kept this at the analytical level rather than reproducing operational detail. The point for defenders is not how to run the attack but what it demonstrates: a mature, revenue-driven access economy is treating perimeter devices as an inventory to be harvested, verified and sold at scale.

Threat landscape: criminal operators, state-grade tooling

Diachenko attributed the operation to a multi-operator, Russian-speaking cybercriminal group, and its behavior. Corporate profiling, resale packaging, an access-broker economy, is consistent with financially motivated crime rather than espionage.

That should not be reassuring. Two factors complicate the picture.

  1. Verified compromises tied to the surrounding activity reportedly include full network intrusions across several countries and, in one case, the exfiltration of classified documents from a NATO defense contractor. That tells us the harvested access is not staying in low-level hands.
  2. Bitsight’s analysis of related exploitation identified state-associated tunneling tools, Chisel and Neo-reGeorg, in the toolkits. The same class of living-off-the-land tunneling infrastructure documented in the Volt Typhoon campaign covered in Briefing #11.

This is the strategic through-line. In Briefing #11 the concern was a state actor quietly pre-positioning inside maritime and adjacent infrastructure by living on edge devices. FortiBleed shows the commercial supply chain that feeds the same terrain: criminal brokers industrialise the harvesting of perimeter credentials, and both ransomware crews and state-aligned operators draw from the resulting pool. An exposed maritime firewall credential is therefore not a single risk but an option that can be exercised by very different actors with very different objectives.

Maritime exposure: what the sector-specific research found

The FortiBleed dataset is global, but the maritime slice of it is substantial and has been examined directly by maritime cybersecurity firm Cydome, whose threat-intelligence team cross-referenced the leaked data against the sector.

Cydome identified more than 250 individual maritime and energy organisations whose Fortinet credentials appear in the known dataset, the majority of them shipowners or ship managers. It also found 703 satellite-linked IP addresses associated with satcom service providers serving maritime operations. A figure that spans both vessels and shore sites, and one that connects this incident directly to the VSAT and satellite-link exposure examined in Briefing #4.

The distribution of affected maritime logins is telling. By Cydome’s breakdown, shipping and freight companies account for 41.5%, offshore contractors and service companies 31.2%, newbuild and repair yards 10.7%, and port authorities and logistics firms 6.7%. In the words of Cydome founder and chief executive Nir Ayalon, the pattern is consistent with FortiBleed reaching the operational core of maritime trade rather than only back-office IT.

Two findings from Cydome’s sampling underline how much exposure remains live. Around 87% of the Fortinet devices examined still had internet-facing management interfaces available, and roughly 63% of the harvested credentials related to default or built-in administrator accounts that had never been renamed. Broader analyses of the full dataset found a similar pattern: a clear majority of compromised accounts were either factory-default or used predictable generic names, meaning attackers could target them with high confidence before spending any effort on brute force. This is not, at root, an exotic technical failure. It is a configuration-hygiene failure repeated across thousands of production deployments.

Article content
The Exposure Was Not Marginal. It Reached the Maritime Core.

Appearing in the dataset is an exposure indicator, not proof of full compromise. But because the harvest date cannot be known, any credential that has not been rotated since the exposure must be treated as live and exploitable. Cydome has published a free lookup at fortibleedcheck.cydome.io so operators can check whether their domain or IP appears. A matching enterprise-facing lookup is also available through Hudson Rock.

Operational impact: from a cracked firewall to the bridge

For a shipping organisation, the operational consequences track the attacker’s path through the network. On the shore side, a compromised perimeter firewall exposes fleet-management platforms, crewing and payroll systems, chartering and procurement, and the Active Directory environment that underpins all of them. The same category of back-office compromise that has driven recent NIS2-reportable incidents in the sector.

On the vessel side, the concern is more acute. Where a FortiGate sits between the satellite [FBB / VSAT / Starlink] link and the onboard network without clean separation between IT and OT, the documented lateral-movement path can reach toward bridge systems, cargo and ballast management, and the satellite communications equipment itself. A firewall turned into a passive listening post can harvest every VPN credential used for remote support of onboard systems. Precisely the remote-access pathways discussed in Briefing #3 and quietly extend the intrusion each time an engineer or vendor connects.

The realistic impact scenarios follow from there:

  • Disruption or manipulation of navigation and cargo systems with direct safety implications
  • Loss of confidentiality over commercially sensitive routing, cargo and chartering data
  • Ransomware deployment staged from the trusted network edge
  • For ports: Disruption to terminal operating systems and logistics flows of the kind examined in Briefing #10.

The financial, regulatory and reputational consequences compound each of these. Critically, none of this requires the attacker to have “hacked” anything in the visible sense. They logged in.

Mitigation: this is remediated by discipline, not a patch

Article content
Operational Discipline Is the Real Fix

Because FortiBleed is a credential and hygiene problem, the response is operational. The following consolidates the guidance issued by CISA and echoed by Fortinet and maritime specialists into a practical maritime sequence.

  1. Check exposure first. Use the available lookups to determine whether your domains and satcom-linked IP ranges [ashore and afloat] appear in the dataset. Absence is not a guarantee of safety, only of absence from the public dataset, so treat it as a starting point rather than an all-clear.
  2. Terminate and rotate. Following CISA’s guidance, terminate all active administrative and VPN sessions and reset every Fortinet VPN and administrator password on internet-facing systems. Do not reuse prior passwords regardless of their complexity, because complexity does not help if the password itself has already been harvested.
  3. Clear the legacy hashes. This is the step most operators miss. Upgrade FortiOS to a current release and ensure every administrator logs in afterward so the credential is re-hashed under PBKDF2. On FortiOS “7.2.x” and “7.4.x”, enable the password-policy setting that removes retained weaker-encryption hashes. Patching alone, without this step, leaves the weak hash in place.
  4. Enforce MFA everywhere. Multi-factor authentication on all administrative and remote-access accounts is the single control that most reliably breaks credential replay, and its widespread absence is a primary reason FortiBleed worked.
  5. Remove management from the public internet. Restrict management interfaces to trusted internal networks or a management VPN, and eliminate internet-facing administration wherever operationally possible. The finding that most affected devices still exposed management interfaces shows how much low-cost risk reduction remains available here.
  6. Hunt, don’t assume. Review logs for unfamiliar administrator logins, unexpected VPN users, configuration changes and new or oddly named accounts, and compare configurations against a known-good baseline. Because compromise is silent, active threat hunting [not the absence of alerts] is what tells you whether access has already been used.

For fleets, the durable lesson reinforces a theme running through this series: enforce genuine segmentation between vessel IT and OT so that a compromised shore-facing or satellite-facing firewall cannot become a route to the bridge, and extend credential rotation and MFA discipline to onboard devices and vendor remote-access accounts, which are too often left on defaults.

Regulatory and compliance considerations

FortiBleed arrives at a moment when maritime cyber accountability is no longer optional and perimeter hygiene is increasingly treated as a compliance issue, not just a technical one.

Under IMO Resolution MSC.428(98), cyber risk must be managed within the vessel’s safety management system, and an unremediated, internet-facing firewall with default credentials is difficult to reconcile with that obligation. The IACS unified requirements UR E26 and E27, applicable to vessels contracted for construction from 1 January 2024, address the cyber resilience of ships and of onboard systems and equipment respectively. E27’s expectations around the integrity and secure configuration of network devices speak directly to the newbuild and repair yards that make up more than a tenth of the affected maritime population, and to the supply-chain assurance concerns raised in Briefing #7. For operators within scope of the EU’s NIS2 Directive, credential exposure of this kind can trigger incident-notification duties and sits squarely within the risk-management and supply-chain security obligations the directive imposes on essential and important entities. In the United States, the Coast Guard’s maritime cybersecurity requirements for the Marine Transportation System place comparable expectations on access control, patching and network segmentation. Flag and port state regimes increasingly test the same ground.

The common thread is that these frameworks no longer treat perimeter hygiene as a technical detail. Default administrator accounts, absent MFA and exposed management interfaces are now compliance failures as much as security ones. And FortiBleed provides an unusually well-documented external record of which organizations were carrying them.

Article content

FortiBleed is not a story about a clever exploit. It is a story about the gap between believing you are secure and being secure.

About upgrades left one step short, default accounts never renamed, and management interfaces left facing the open internet. For a sector as connected and as edge-dependent as shipping, that gap is expensive, because the firewall FortiBleed targets is often the last line between a satellite link and a ship’s bridge.

The maritime findings should be read as a prompt, not a verdict. More than 250 maritime and energy organisations in the dataset, hundreds of satcom-linked addresses, and a clear majority of exposure traceable to default credentials all point to a problem that is broad but also, encouragingly, largely remediable with disciplined action already within every operator’s reach. The uncomfortable part is that appearing [or not appearing] in a public dataset tells you very little about whether your own keys have already been copied. The only reliable answer comes from rotating credentials, clearing legacy hashes, enforcing MFA, pulling management off the internet, and hunting for access that may already have been used.

The resilient operators in twelve months’ time will not be the ones who patched fastest in July 2026. They will be the ones who treated their perimeter as a living surface requiring continuous verification, and who assumed, until proven otherwise, that the keys were already out.


Maritime Cyber Threat Briefing is an independent series covering cyber threats, vulnerabilities, and risk management across the global maritime industry. It is published by Alexandros Engelen, a Cybersecurity Strategist, specializing in maritime cyber risk.

Sourcing disclosure: This briefing draws on the original FortiBleed disclosure by Volodymyr Diachenko and verification by Kevin Beaumont (DoublePulsar) and Hudson Rock; technical analyses by SOCRadar, Arctic Wolf, Recorded Future (Insikt Group), Bitsight and the Cloud Security Alliance; Fortinet’s PSIRT analysis and CISA guidance. And maritime-specific research published by Cydome. Maritime figures, the satcom-linked exposure count and the sector breakdown are as reported by Cydome. Attribution and victim details reflect the assessments of the cited researchers and firms; appearance in the dataset is an exposure indicator, not confirmation of compromise. No confidential or non-public information was used.

When the Shore Side Falls: Dissecting the Anubis Ransomware Attack on the Adriatic Port Authority

Maritime Threat Craft #2

The Anubis ransomware group’s attack on the Adriatic Port Authority, the “Autorità di Sistema Portuale del Mare Adriatico Centrale“ [operator of the Port of Ancona], was not a sophisticated OT intrusion. No ECDIS was touched. No ballast management system was manipulated. No integrated bridge system received a command it shouldn’t have. The publicly reported activity [attackers] remained within shore-side IT infrastructure:

  • Office 365 accounts
  • Azure management layers
  • Windows endpoints
  • and whatever unpatched internet-facing services happened to be in scope.

And yet the operational consequences were real and immediate. Cargo processing halted, vessels rerouted, regional supply chain throughput disrupted.

That is the threat model maritime security practitioners need to internalize. The attack surface that matters is not just the OT stack on the vessel. It is the digital scaffolding that makes a port function: the logistics platforms, customs processing systems, shipping schedule databases, and workforce management applications that sit entirely in IT but whose failure cascades directly into the physical domain. Understanding how an adversary thinks about that scaffolding [and how they exploit it] is the purpose of this analysis.

Article content
A conventional ransomware chain becomes maritime disruption when shore-side trust systems fail.

What made this maritime?

The maritime relevance is not the malware family. It is the dependency chain. Anubis did not need bridge-system access, crane-control access, or vessel OT access to create maritime consequence. The leverage sat in shore-side trust functions: cargo status, customs workflow, berth and schedule coordination, employee access, and operational documentation. Those systems are administratively classified as IT, but operationally they behave as maritime control points.


Anubis is a financially motivated ransomware-as-a-service (RaaS) operation. Its objectives are not strategic in the nation-state sense. The primary objective was monetization, not strategic pre-positioning. But the stolen dataset can retain intelligence value long after the extortion event ends. The objective is monetization, structured around a double extortion model: encrypt systems to deny access, exfiltrate data to maximize pressure, demand payment under a timed deadline, and follow through on data release if refused.

The group emerged in December 2024, publicly launched its affiliate program on the RAMP (Russian Anonymous Marketplace) cybercrime forum in February 2025, and operates under the alias “superSonic” on RAMP and “Anubis__media” on XSS and Exploit underground forums. By the time it listed the Adriatic Port Authority on its data leak site in January 2026, the group claimed revenues exceeding $20 million. A figure consistent with a mid-tier RaaS operation running at scale across multiple verticals.

What sets Anubis apart within the current RaaS ecosystem is its tiered affiliate compensation structure. Rather than a standard revenue split, the group advertises differentiated cuts: 80% to affiliates deploying ransomware, 60% for data-only extortion operations, and 50% for initial access brokers (IABs) providing network footholds without executing the payload. This structure explicitly incentivizes IABs to bring maritime and critical infrastructure targets into scope, because the group actively purchases stolen corporate data of 40 GB or more from enterprises in the US, Canada, EU, and Australia. Provided the data has not previously appeared in open sources.

The port authority as a target serves the financial objective in a specific way. Maritime operators are under intense operational pressure: cargo dwell time, port call turnaround, customs clearance windows. All of these are measured in hours, not days. When core logistics systems become inaccessible, the pressure to pay accelerates faster than in a typical enterprise environment. The adversary understands this. The $10 million Bitcoin ransom demand, delivered with a seven-day countdown and a threat to publish stolen data, was calibrated for that pressure.

There is a secondary dimension worth noting for the practitioner: the data Anubis extracted from the port authority included safety plans and information about security operations. Resecurity, the threat intelligence firm whose analysis anchored reporting on this incident, explicitly flagged that this category of data carries value far beyond the ransomware context. Ιt is precisely the kind of intelligence that organized crime groups involved in smuggling, contraband, and insider recruitment would pay to access. The ransomware attack may have had a purely financial objective, but the stolen dataset has a longer tail.


Target Selection Logic

Article content
Anubis valued reachability, pressure, and leverage — not maritime complexity.

The Adriatic Port Authority was not targeted because an Anubis affiliate had specialized knowledge of maritime infrastructure. It was targeted because it exhibited the profile characteristics that Anubis [and the RaaS ecosystem] broadly uses to identify accessible, high-pressure targets.

  1. Digitalization without maturity. Port authorities have undergone rapid digitalization over the past decade. Cargo tracking, customs processing, vessel scheduling, and workforce systems are now entirely platform-dependent. But the pace of digitalization has consistently outrun the pace of security investment. Resecurity’s assessment is direct: IT systems in the port segment are “extremely outdated and lack cybersecurity maturity,” and most ports remain unprepared for large-scale systemic cyber incidents. This is not an assessment unique to the Port of Ancona, it describes the sector condition.
  2. Internet-facing surface exposure. Anubis affiliates conduct reconnaissance against internet-facing services before committing to a target. The group’s documented initial access vectors include SonicWall VPN appliances operating without MFA, Cisco SSL VPN endpoints, SolarWinds Web Help Desk instances vulnerable to CVE-2025-26399, and CitrixBleed 2 (CVE-2025-5777). Any organisation running these systems without current patch levels and MFA enforcement is visible from Shodan and Censys scans. Port authority IT infrastructure [which typically includes remote management portals, vendor connectivity gateways, and cloud management interfaces] creates significant external exposure that is trivially enumerable by an IAB doing pre-targeting reconnaissance.
  3. Cloud misconfiguration as an attack surface. The Adriatic attack was executed through weaknesses in Office 365 and Azure account management. This is a consistent pattern in modern ransomware operations: cloud identity stores that lack conditional access policies, legacy authentication protocols still enabled, service accounts with excessive permissions, and MFA gaps in hybrid environments create pathways that bypass traditional perimeter controls. Port authorities that migrated to cloud productivity platforms without correspondingly maturing their identity security posture created exactly this exposure.
  4. Workforce as the weakest node. The affiliate responsible for initial access specifically targeted employees of the company managing the port authority not port operations staff directly, but the administrative and management workforce with privileged access to production systems. From the adversary’s targeting logic, this makes sense. Shore-side administrative staff typically have broader Active Directory privileges than operational staff, operate on general-purpose Windows endpoints with email and browser exposure, and receive security training that is less rigorous than vessel crew training (itself often inadequate). The human attack surface in the port management environment is large and soft.
  5. Geopolitical and geographic context. The Port of Ancona sits on Italy’s central Adriatic coast and serves as a primary gateway for cargo and passenger traffic between Italy and the Western Balkans [Albania, Greece, Croatia, Montenegro]. It handles approximately 10 million tonnes of cargo annually and is a significant node for Ro-Ro, container, and passenger ferry operations. This geographic and commercial position means that disruption at Ancona has cascading effects across the Adriatic corridor. An adversary [criminal or state-aligned] targeting maximum regional economic disruption would identify Ancona as a high-leverage node. Resecurity’s observation that nation-state actors could use identical TTPs for gray-zone operations is not speculative. It reflects a documented convergence between RaaS tradecraft and hybrid warfare methodology that practitioners need to track.

Access Sequencing

The initial access vector was a spear-phishing email containing a malicious attachment, delivered to employees of the company managing the port authority. This is consistent with Anubis’s documented methodology and with the broader affiliate-sourced access model.

Understanding the access sequencing in this attack requires separating three phases: initial foothold establishment, credential harvesting and privilege escalation, and lateral movement to target systems.

Phase 1 – Initial foothold. The spear-phishing email exploited the most consistently reliable initial access vector in the sector: the human endpoint. In a port authority environment, the employees most likely to be targeted are those with visible roles: procurement officers, port captain administrative staff, HR and payroll administrators, and IT helpdesk personnel. These roles are often identifiable through publicly available port authority websites, shipping industry directories, LinkedIn profiles, and regulatory filings. An affiliate with even basic OSINT capability can construct a plausible pretexting scenario [a shipping agent inquiry, a regulatory compliance request, a payroll discrepancy notice] that warrants opening an attachment. The Adriatic port authority’s workforce represents a broad and accessible spear-phishing surface.

The malicious attachment’s execution would have established an initial foothold within the Windows environment. Anubis’s documented toolkit includes capabilities consistent with a staged loader architecture: the initial payload establishes persistence and beaconing, then pulls additional tooling based on environment reconnaissance. Trend Micro‘s binary analysis identified the payload as employing ECIES (Elliptic Curve Integrated Encryption Scheme) for file encryption, implemented in Go using a publicly available ECIES library. The same library and implementation pattern appears in the EvilByte/Prince ransomware family, which is a useful data point for practitioners building YARA or behavioral signatures. This is shared tooling lineage, not a bespoke cryptographic implementation, and detection content built against one family has a reasonable chance of generalizing to the other.

The practical defender consequence of ECIES specifically [as opposed to the AES-based symmetric schemes still common across the ransomware landscape] is that there is no shared-secret weakness to exploit. AES-based ransomware occasionally yields to key-recovery research when implementation errors produce reusable keystreams or weak per-file IVs. ECIES binds the decryption capability to an asymmetric keypair the operator controls outside the victim environment; absent an implementation flaw or key exposure, there is no realistic recovery path without the operator-controlled private key. For incident response planning, this means: budget for the assumption that payment is the only path to file recovery if backups fail, and weight your investment accordingly toward backup immutability rather than toward the (low) probability of a public decryptor appearing.

Phase 2 – Credential harvesting and escalation. From the initial foothold, the affiliate moved to credential harvesting. In an Office 365 and Azure-integrated environment, this phase is particularly consequential. Once an adversary has a foothold on a domain-joined endpoint with cached credentials, standard post-exploitation tooling can extract LSASS memory for NTLM hashes, query the Windows Credential Manager for stored credentials, and enumerate Microsoft Entra ID (formerly Azure Active Directory) tenant configuration for service principals, app registrations, and federated identity providers. If the compromised account held even Global Reader permissions, the attacker could enumerate the entire tenant’s role assignments and identify the shortest path to a Global Administrator or Privileged Role Administrator account.

Article content
One compromised identity can turn port business data and cloud control into the same attack surface.

This is the structural failure this diagram is built to surface: in many shore-side port authority IT environments, there is no segmentation between the credential set that manages day-to-day productivity (mailboxes, document libraries, HR records) and the credential set that manages cloud infrastructure (virtual machines, storage accounts, app registrations). A single compromised identity with broad role assignment can pivot in either direction. Resecurity’s finding that the Adriatic intrusion exploited “insecure accounts managing Office 365/Azure” is consistent with exactly this flat-privilege condition. The same account that gave the attacker access to contracts and employee records likely also gave them visibility into broader tenant infrastructure, even though the documented impact stayed within the data and endpoint domain.

QEMU-based VM abuse has appeared in current ransomware tradecraft adjacent to this ecosystem, and it belongs in the hunting model for port environments even if it is not publicly confirmed as part of the Adriatic intrusion. Deploying a lightweight QEMU virtual machine on a compromised host creates an isolated execution environment that is invisible to endpoint detection and response (EDR) tools operating at the host OS level. The attack tooling [lateral movement utilities, credential harvesters, encryption preparation stages] can operate inside the VM without triggering EDR telemetry on the host. This technique, independently documented by Sophos researchers in the context of related RaaS campaigns leveraging SonicWall vulnerabilities, represents a meaningful evasion capability that degrades detection probability in environments running standard EDR products without network-level behavioral analytics.

The mechanics worth understanding: QEMU’s system emulator binary [qemu-system-x86_64.exe on Windows] can be dropped onto a host and used to boot a minimal Linux guest image in seconds. From inside that guest, the attacker has a fully functional, internet-routable execution environment that shares the host’s network adapter via a bridged or NAT configuration but presents no process tree, no file writes, and no API call surface to host-based EDR. Because as far as the hypervisor’s host process is concerned, the only activity is a single long-running QEMU process making routine memory and disk I/O calls. Tooling executed inside the guest [credential dumpers, SMB enumeration scripts, exfiltration utilities] generates network traffic that looks, to the host, identical to ordinary QEMU disk-image I/O.

Phase 3 – Lateral movement to target systems. With escalated credentials, the attacker moved laterally across the port authority’s IT infrastructure to reach systems managing cargo tracking, shipping schedules, and customs processing. In a port authority environment, these systems are likely to include Terminal Operating System (TOS) components, port community system (PCS) interfaces, and customs messaging gateways. All of which are functionally IT applications, not OT systems, but whose operation has direct physical-domain consequences. The attack did not need to touch any OT or maritime-specific technology stack. The IT targeting was sufficient.

The fact that insecure accounts managing Office 365 and Azure were the documented exploitation pathway suggests that the lateral movement phase leveraged cloud management credentials harvested from the initial foothold. Likely a Global Admin or privileged role account whose credentials were cached or reused across endpoints. This is a consistent pattern: shore-side port authority IT environments often have flat privilege structures where administrators manage both operational applications and cloud tenancy from the same credential set.


Persistence and Dwell

The initial compromise occurred on December 11, 2025. Attribution to Anubis was established publicly in January 2026, a gap of approximately four weeks. The data leak site listing appeared in mid-January. This timeline implies a dwell period of several weeks between initial access and encryption deployment.

In a shore-side IT environment, dwell period tactics are more conventional than in a vessel OT environment. The adversary does not face the connectivity constraints, protocol limitations, or air-gap architectures that complicate persistence in shipboard systems. What dwell looks like in this environment:

  1. Persistence mechanisms. Anubis employs valid account usage as its primary persistence mechanism. A deliberate choice that minimizes artifact creation. Using harvested credentials to authenticate normally to systems generates authentication logs that blend with legitimate traffic. In an environment without behavioral analytics capable of detecting authentication anomalies (unusual access times, source IP changes, atypical resource access patterns), this persistence is essentially invisible to routine log review. Scheduled task creation and Windows service registration have also been observed in Anubis-adjacent campaigns, providing execution persistence that survives credential rotation if defenders do not identify and purge all access vectors.
  2. Reconnaissance during dwell. The dwell period served a reconnaissance function: mapping the environment, identifying backup systems and their connectivity, locating data repositories of highest extortion value, and staging exfiltration before encryption. The safety plans and security operations data that appeared in the leaked dataset required deliberate identification and targeting during this phase. The adversary had to navigate the file system and network shares to find this material. That navigation takes time and generates telemetry that an instrumented environment would detect.
  3. Pre-encryption preparation. Before deploying the encryption payload, Anubis affiliates typically attempt to disable backup connectivity, terminate security processes, and clear event logs. Anubis samples analyzed by Trend Micro document a privilege check gate, the binary verifies administrative privileges before executing its most destructive functions, followed by a specific shadow-copy purge command:
vssadmin delete shadows /for=norealvolume /all /quiet

This is worth committing to memory as a detection primitive in its own right. The “/for=norealvolume” flag is an unusual construction, most legitimate administrative shadow-copy cleanup targets a specific drive letter or volume GUID, not a placeholder non-volume. Any execution of vssadmin delete shadows with this exact flag pattern, or any shadow-copy deletion command invoked from a process spawned by an Office document handler [winword.exe, excel.exe, outlook.exe as a parent process], should be treated as a near-certain ransomware precursor rather than routine maintenance. In the Adriatic environment, the reliance on “outdated backup protocols” that slowed recovery suggests that backups were both shadow-copy dependent and reachable from the attacker’s session. An environment with properly segmented, immutable, offline backups would have a materially different recovery timeline regardless of whether shadow copies were purged.

One critical detail worth flagging: Anubis has documented wiper functionality that can be operator-activated. Trend Micro’s analysis identified command-line options within sampled binaries that, when invoked, cause the ransomware to permanently destroy file contents rather than encrypt them, making recovery impossible even after payment. In the Adriatic incident, this capability was apparently not activated. The port authority was able to restore from backups, and Resecurity described the impact as approximately 2% data loss. But the capability exists within the toolkit. Any port authority that assumes ransomware attacks are always recoverable from backups needs to account for the possibility of deliberate wiper activation by an affiliate motivated to maximize impact rather than maximize payment probability.


Objective Execution

Article content
Anubis executed for pressure: exfiltrate first, encrypt second, force payment through operational disruption.

The encryption event was the operational culmination of the dwell and preparation phases. When the payload executed, it propagated across the accessible network, encrypting files on endpoints and network shares using ECIES-based encryption. Systems supporting cargo tracking, shipping schedules, and customs processing became inaccessible. The port authority’s ability to process incoming and outgoing shipments collapsed.

The operational consequences were immediate and physical:

Vessels in the approach phase for Ancona were rerouted to alternative ports likely Ravenna, Bari, or Brindisi depending on cargo type and commercial arrangements. Rerouting a vessel in the Adriatic is not a complex navigation problem, but it is commercially costly. Repositioning fees, demurrage charges, revised berthing schedules, and consignee notification across the cargo manifest. For Ro-Ro and passenger ferry operators on the Ancona-Igoumenitsa, Ancona-Patras, and Ancona-Split routes, disruption extends beyond freight to passenger operations and vehicle logistics.

The seven-day payment countdown imposed by the ransom note was calibrated against this operational pressure. The adversary’s timing logic is straightforward. A port authority facing daily losses from halted cargo processing, vessel rerouting costs, and contractual penalties accumulates financial pressure that compounds faster than the recovery timeline for a system encrypted at scale. The $10 million demand represents a fraction of the projected operational losses if the disruption persisted for weeks.

The secondary objective [data publication as leverage] functions differently. The leaked dataset included contracts, employee records, and critically, safety plans and security operations documentation. Employee records create regulatory exposure under GDPR for an Italian public entity: the port authority faces potential enforcement action by the Garante Privacy in addition to operational recovery. Safety plans and security operations documentation, once circulated on Anubis’s Tor-hosted leak site (om6q4a6cyipxvt7ioudxt24cw4oqu4yodmqzl25mqd2hgllymrgu4aqd.onion), become permanently accessible to any actor with capability and intent. Including criminal networks operating smuggling routes through the Adriatic corridor, where the Port of Ancona is a primary transit point.

This is the dimension of the attack that matters beyond the immediate incident. Ransomware is increasingly the delivery mechanism for intelligence collection that has utility far beyond the extortion event itself. The operational security plans of a port authority are not just embarrassing if leaked, they are operationally useful to criminal and state-aligned actors who need to understand the physical security posture of the facility, checkpoint locations, shift patterns, and vessel screening procedures.


Defender Implications

Model port-community trust explicitly. A port authority compromise does not remain inside the authority if its systems exchange data with shipping agents, terminal operators, customs brokers, ferry companies, logistics providers, vessel operators, or public-sector agencies. Defenders should inventory which external parties can submit, modify, approve, or consume operational data through PCS, TOS, email, API, SFTP, or portal access. The question is not only “what systems were encrypted?” It is “which maritime decisions depended on data from those systems?”

The Anubis/Adriatic incident is instructive precisely because the attack did not require OT targeting, vessel-specific tradecraft, or maritime technology expertise. The adversary achieved physical-domain consequences through entirely conventional IT exploitation. This has specific implications for maritime security practitioners.

Reconceptualize the shore-side IT environment as maritime critical infrastructure. Port authority IT systems managing TOS, PCS, and customs interfaces are not enterprise IT in the generic sense. Their failure has consequences measured in vessel diversions, cargo demurrage, and supply chain disruption. They require the same asset criticality classification and network segmentation posture that you would apply to OT systems. Practitioners managing port security architectures should map every IT dependency that, if unavailable, generates operational consequences in the physical domain. That map defines your actual critical asset boundary.

Threat hunt specifically for QEMU deployment. The documented use of QEMU virtual machine abuse for defense evasion means that host-based EDR telemetry alone is insufficient. Detection requires network behavioral analytics capable of identifying anomalous east-west traffic patterns from hosts running unexpected hypervisor processes. A starting detection rule, in Sigma-style logic, adaptable to your SIEM:

title: Unexpected QEMU system emulator execution
logsource:
  category: process_creation
  product: windows
detection:
  selection:
    Image|endswith:
      - '\qemu-system-x86_64.exe'
      - '\qemu-img.exe'
  filter_known_virt_hosts:
    ComputerName|in:
      - <list of approved virtualization/lab hosts>
  condition: selection and not filter_known_virt_hosts
level: high

Pair this with a network-layer correlation: alert on any host generating sustained outbound connections where the parent process is “qemu-system-x86_64.exe” and the host is not in your approved hypervisor inventory. This technique is not yet widely detected by default EDR configurations, which is precisely why it appears in current RaaS tooling. Build the rule before you need it, not after.

Hunt for the shadow-copy deletion precursor. The specific command construction documented in Anubis samples is distinctive enough to alert on directly:

title: Anubis-style shadow copy deletion via vssadmin
logsource:
  category: process_creation
  product: windows
detection:
  selection:
    Image|endswith: '\vssadmin.exe'
    CommandLine|contains|all:
      - 'delete'
      - 'shadows'
      - '/for=norealvolume'
  condition: selection
level: critical

Layer a second, broader rule beneath it that catches the technique family rather than this one variant. Any vssadmin delete shadows /all /quiet invocation, any wmic shadowcopy delete, or any wbadmin delete catalog execution, correlated against parent process. If the parent is an Office application, a script host (wscript.exe, cscript.exe, powershell.exe), or a process with no prior baseline of administrative tool usage, escalate immediately, don’t wait for the encryption event to confirm the alert.

Treat cloud identity posture as your primary attack surface. The Adriatic attack succeeded through cloud account compromise. Office 365 and Azure management interfaces operated with insufficient access controls, consistent with the flat-privilege escalation path shown above. Practitioners managing maritime IT environments running Microsoft 365 should audit:

  • Conditional access policy coverage for all privileged roles. No exceptions for “legacy” service accounts.
  • Legacy authentication protocol status: SMTP AUTH, IMAP, POP. All should be disabled unless operationally required and compensated with additional controls.
  • Global Administrator and Privileged Role Administrator assignments. Reduce to minimum viable count, enforce phishing-resistant MFA.
  • Segregation between identities that hold Exchange/SharePoint data access and identities that hold Azure management plane [subscription, resource group, app registration] permissions. No single account should span both domains without break-glass procedures and just-in-time elevation.

Establish backup isolation as a non-negotiable architecture requirement. The Adriatic recovery was delayed because backup protocols were outdated and apparently accessible from the compromised environment. Backups connected to the domain [reachable via SMB shares authenticated with domain credentials] are not backups in any meaningful ransomware-resilience sense. Maritime IT environments must implement immutable backup targets [WORM storage, air-gapped tape, or cloud storage with object lock policies] and test restoration quarterly under timed conditions. Know your actual RTO for each critical system before the incident, not during it.

Build detection logic for the specific exfiltration window. In the Adriatic incident, the four-week dwell period before encryption included a deliberate exfiltration phase targeting safety plans and operational security documentation. This exfiltration generates detectable telemetry: large volume transfers from file servers to endpoints or external destinations, access to directories containing document classes not normally accessed by the account in question, and credential-authenticated access to SharePoint sites or OneDrive repositories outside normal working patterns. A starting query skeleton, in KQL form for Microsoft 365 / Sentinel environments:

OfficeActivity
| where Operation in ("FileDownloaded", "FileSyncDownloadedFull", "FileAccessed")
| where ObjectId has_any ("security", "safety-plan", "ISPS", "contingency", "operations")
| summarize FileCount = dcount(ObjectId), TotalEvents = count()
    by UserId, bin(TimeGenerated, 1h)
| where FileCount > 15 or TotalEvents > 50
| join kind=leftanti (
    OfficeActivity
    | where TimeGenerated < ago(30d)
    | summarize by UserId
) on UserId

The join against a 30-day historical baseline is the key element. It isolates accounts that are suddenly accessing sensitive document classes at volume with no prior access history, which is a stronger signal than volume alone. DLP policies scoped to sensitive document types [port security plans, operational procedures, personnel records] with alerting on bulk access events would create detection opportunity during the dwell phase. Before the encryption event, and before the exfiltrated material has any chance of reaching a leak site.

Actively threat hunt for IAB activity. Anubis’s affiliate program structure means that initial access may be sold to a broker who does not immediately deploy the payload. Between IAB access and affiliate payload deployment, there is a window [potentially weeks] during which the adversary is present but inactive. Hunt for indicators of low-and-slow reconnaissance: LDAP enumeration queries from unusual source accounts, WMI or PowerShell remote execution from endpoints not normally used for remote management, authentication patterns showing off-hours access to document repositories, and new scheduled task registrations that do not correspond to known change tickets.

Understand the intelligence value of your own data. The port authority’s safety plans and security operations documentation reached Anubis’s leak site. For maritime practitioners: classify your data by its downstream intelligence value to criminal and state-aligned actors, not just by its regulatory sensitivity. Cargo manifests, vessel arrival and departure schedules, berth allocation records, physical security overlays, and customs processing workflows all have intelligence value beyond their operational function. That classification should drive access control decisions.

Track the threat escalation trajectory. Resecurity projects that attacks against port authorities and maritime operators will intensify markedly through 2030. Driven by geopolitical tensions, expanding digital connectivity in vessel and port environments, and the maturation of the RaaS affiliate ecosystem. The IACS UR E26 and E27 cybersecurity requirements impose baseline security standards on newbuild vessels from 2024, but shore-side port infrastructure is governed by a patchwork of national regulations with inconsistent enforcement. The attack surface for RaaS affiliates and the nation-state actors who observe their TTPs closely is widening, not narrowing.


MITRE ATT&CK Mapping (Selected Techniques – Adriatic Incident)

Article content
The Adriatic incident mapped conventional IT compromise to maritime operational consequence.
  • T1566.001 – Spearphishing Attachment – Initial Access
  • T1190 – Exploit Public-Facing Application – Initial Access (IAB sourcing)
  • T1078 – Valid Accounts – Persistence / Defense Evasion
  • T1068 – Exploitation for Privilege Escalation – Privilege Escalation
  • T1003.001 – OS Credential Dumping: LSASS Memory – Credential Access
  • T1021.002 – Remote Services: SMB/Windows Admin Shares – Lateral Movement
  • T1562.001 – Impair Defenses: Disable or Modify Tools – Defense Evasion
  • T1564.006 – Hide Artifacts: Run Virtual Instance (QEMU) – Defense Evasion
  • T1005 – Data from Local System – Collection / Exfiltration
  • TT1485 – Data Destruction – Impact
  • T1486 – Data Encrypted for Impact – Impact
  • T1490 – Inhibit System Recovery (shadow copy deletion) – Impact
  • T1491 – Defacement / Ransom Note Delivery – Impact

Incident Timeline

Article content
The Adriatic incident unfolded as a conventional IT intrusion with maritime operational consequences.
  • December 11, 2025 – Initial compromise – spear-phishing delivers foothold
  • December 2025 (ongoing) – Dwell period – reconnaissance, credential harvest, exfiltration of safety plans and employee records
  • December 2025 / January 2026 – Encryption payload deployed – cargo systems, scheduling, customs processing rendered inaccessible
  • January 13–14, 2026 – Anubis lists Adriatic Port Authority on data leak site; $10M Bitcoin ransom demand with seven-day deadline
  • January 2026 – Breach publicly attributed; port authority issues statement; incident response engaged
  • June 11, 2026 – Resecurity publishes detailed threat intelligence analysis of the incident

Technical Appendix: Anubis Infrastructure and Negotiation Mechanics

For practitioners building threat intelligence profiles, the operational backend behind the Adriatic incident has several documented characteristics worth tracking.

Forum presence and attribution signals. Anubis operators maintain parallel personas across two Russian-language cybercrime forums: “superSonic” on RAMP (Russian Anonymous Marketplace) and “Anubis__media” on XSS and Exploit. Both accounts post exclusively in Russian. On February 23, 2025, “superSonic” advertised a restructured affiliate program on RAMP, explicitly noting that all revenue-share figures were negotiable for long-term cooperation. A recruitment posture aimed at retaining skilled affiliates rather than running a fixed-rate marketplace. Negotiation activity, leak-site updates, and hands-on attack operations cluster around Moscow Standard Time business hours, and ransom notes and binaries contain Russian-language strings and occasional Cyrillic characters left uncleaned in the code. The standard combination of indicators threat intelligence teams use to assess CIS-region operator origin, though this remains an assessment rather than confirmed attribution.

Targeting exclusions. Anubis has publicly stated it excludes CIS-region states, educational institutions, government bodies, and non-profit organizations from targeting [a now] standard self-imposed restriction across Russian-speaking RaaS operations, generally read as an operational security measure to avoid domestic law enforcement attention rather than an ethical constraint. Port authorities are quasi-governmental in many jurisdictions; the Adriatic Port Authority’s status as an Italian public entity did not trigger this exclusion, which tells practitioners that the exclusion list is narrower than “anything government-adjacent”. It is specifically calibrated to avoid direct domestic-state friction, not to avoid critical infrastructure generally.

The three-track affiliate economy. The structure is more granular than a simple ransomware RaaS: a traditional deployment track (80% to affiliate, 20% to Anubis), a data-only extortion track for affiliates who already hold stolen data and want monetization-as-a-service (Anubis requires the data be under six months old and not previously published, in exchange for a share of extortion proceeds), and an access-resale track for initial access brokers who sell footholds without further involvement. This tripartite structure is precisely why a single ransomware “group” can appear behind incidents with very different technical signatures. The hands-on intrusion specialists, the data monetization specialists, and the access brokers may never be the same individuals, and an IAB who sold access to the Adriatic Port Authority’s network may have had no further involvement in the encryption event or negotiation.

Lead generation as published content. Anubis prepares a detailed, investigative-article-style write-up for each victim, built from the exfiltrated data, before publication. This write-up is initially posted to the group’s Tor-hosted leak site in hidden mode. Accessible only via a direct link shared privately with the victim during ransom negotiations, functioning as proof of compromise and a pressure tool simultaneously. If negotiations fail or the deadline lapses, the article [and the underlying file tree] is made publicly downloadable. The ransom note dropped on encrypted systems is a static HTML file named RESTORE FILES.html. Victim communication during negotiation has been documented via the email aliases “anubis@mailum[.]com” and “anubis20@firemail[.]de”, though the group represents itself as willing to negotiate data removal from the leak site even after publication. A claim threat intelligence analysts should treat as a negotiation tactic rather than a guarantee, since data removed from a leak site has no verifiable assurance against prior copying or onward sale.

Leak site operational status. The group’s Tor hidden service [catalogued by RansomLook at the address listed in the sources below] has maintained roughly 87% uptime over rolling 30-day windows, posts new victims at a rate of approximately ten to fifteen per month, and is protected by a CAPTCHA gate. Standard infrastructure hardening against automated scraping by takedown services and researchers. Practitioners running dark web monitoring programs should track posting cadence as a leading indicator: a sudden increase in posting frequency from a tracked group often precedes or accompanies an affiliate recruitment push, which in turn precedes a wave of opportunistic targeting against whatever internet-facing vulnerability the new affiliate cohort has been equipped to exploit.


The Adriatic incident does not exist in isolation. The maritime port sector has absorbed a series of ransomware attacks over the past decade that collectively define the threat pattern:

Article content
Across a decade of port ransomware, the pattern is consistent: conventional IT compromise becomes maritime operational disruption.
  • NotPetya’s 2017 destruction of Maersk’s global IT infrastructure [including its terminal operating systems] resulted in $200-300 million in losses and required a ten-day recovery operation involving reinstallation of approximately 45,000 PCs and 4,000 servers. That attack was nation-state attributed (Russia/Sandworm), demonstrated the physical consequences of maritime IT disruption at global scale, and served as the proof-of-concept that the RaaS ecosystem subsequently adopted for port targeting.
  • LockBit 3.0’s 2023 attack on the Port of Nagoya’s NUTS (Nagoya United Terminal System) halted container cargo operations for two days and disrupted Toyota’s supply chain.
  • A 2026 attack on the Port of Vigo, Spain [operator unknown at time of writing] forced manual cargo operations as digital systems failed.

The pattern: maritime ports are consistently accessible through conventional IT attack vectors, consistently lack mature detection and response capabilities, and consistently face operational pressure that accelerates payment decisions. Anubis affiliates operating against the Adriatic Port Authority were executing a playbook that the threat intelligence community has documented across dozens of incidents. The TTPs were not novel. The targeting logic did not need to be sophisticated. It only needed to align with how port operations actually depend on digital trust.

That effectiveness is the practitioner’s problem to solve.


Sources and Further Reading

Maritime Threat Craft is an independent LinkedIn article series written for maritime IT/OT security practitioners. Series identity: Adversary Thinking for Maritime Defenders. It is published by Alexandros Engelen, a Cybersecurity Strategist, specializing in Maritime Cyber Risk.

How an Adversary Thinks About a Vessel as a Target

Maritime Threat Craft #1

For five months, an adversary sat inside the management infrastructure of an Iranian satellite communications provider and did nothing visible. No ransom note, no defacement, no operational disruption. They mapped a fleet [modem by modem] across the estates of two of Iran’s largest shipping operators. Then, in two timed waves in March and August 2025, Lab Dookhtegan (meaning “Sewn Lips” in Farsi) severed communications across the NITC and IRISL fleets [116 vessels in the March wave alone] by destroying the Falcon stack on every connected terminal at once. They never touched a ship.

That operation is a useful place to start, because it inverts the assumption most maritime security programs are built on: that the vessel is the thing you defend. The adversary in that campaign never regarded the vessel as the target. The vessel was downstream of the target. What they attacked was the management plane that made the fleet’s connectivity coherent and from the defender’s seat aboard any one of those 116 ships, there was nothing to see until the moment of execution.

Article content
When the management plane is the target, the vessel becomes the blast radius.

This is the gap this series exists to close. Maritime security is dominated by compliance frameworks that describe what to protect. They are largely silent on how an adversary decides whether you are worth attacking, where they will enter, and how they will think about you once they are inside. Those are different questions, and they have different answers depending on who is asking.

Before any adversary moves, they answer one thing: why this target, and what does success look like? Maritime assets complicate that question in ways most IT/OT environments do not. A vessel is simultaneously a mobile node, a logistics platform, a communications relay, a cargo container, and for some operators an instrument of state. An adversary assessing a ship is not assessing a building with a network inside it. They are assessing a system that moves, connects intermittently, carries different things at different times, and sits at the intersection of multiple jurisdictions, none of which holds clean investigative authority.

That complexity is not a problem for adversaries. It is a feature.

And the objective determines everything downstream of it. Nation-state actors in the maritime domain:

  • Mustang Panda against European cargo operators
  • APT28 surveilling NATO logistics chains
  • Crimson Sandstorm probing Mediterranean shipping

They are not running the playbook of criminal ransomware affiliates, and neither category maps onto hacktivist collectives like Lab Dookhtegan. Three adversary classes, three objective sets, three fundamentally different ways of evaluating the same ship. A defender who collapses them into a single “threat” has already miscalibrated. The question is never “could someone attack us?” It is “which adversary class would mark us as a valid target, and what would success mean to them?”


Target Selection Logic

Adversaries do not choose vessels at random. They run a target valuation process [often implicit, sometimes quite deliberate] that weighs return on effort against probability of success and risk of exposure. Understanding that valuation process is the foundation of useful threat modelling.

  1. Vessel class and cargo type are the first filter. Tankers carrying sanctioned oil or liquefied natural gas (LNG) are high-value to hacktivist actors pursuing disruption objectives and to nation-state actors mapping energy dependency. Container vessels operating on trans-Pacific or Asia-Europe routes attract intelligence collection interest. Cargo manifests and commercial shipping data have strategic value when aggregated. Bulk carriers operating under flags of convenience are attractive to criminal actors because operational disruption pressure is high, legal recourse is slow, and crew capacity to respond to a cyber incident is typically low.
  2. Flag state and regulatory posture matter more than defenders assume. An open-registry vessel with minimal ISM Code cyber enforcement carries a different risk profile than one under a flag with active port state control, and OSINT surfaces the distinction quickly. Flag also shapes the post-incident environment: jurisdictional ambiguity is useful to any adversary working to stay below the attribution threshold.
  3. Trade route and geopolitical exposure are the next filter, chiefly for nation-state and hacktivist actors. A vessel regularly transiting Hormuz, the Red Sea, or the Taiwan Strait sits in active friction zones where digital targeting extends a hybrid campaign. APT28 is documented against NATO maritime supply chains in the Baltic; Chinese-nexus actors surveil vessels whose routes cross South China Sea claims; Lab Dookhtegan selected NITC and IRISL ships not for their posture but because they served Iranian state economic objectives and were already sanctioned “politically legitimate targets”, in the group’s framing.
  4. Electronic warfare as target manipulation has matured from opportunistic tactic to systematic coercion tool. GPS spoofing and jamming in sensitive chokepoints [the Red Sea, the Persian Gulf and Strait of Hormuz, the Baltic, and the Black Sea] are no longer anomalous. In June 2025, Windward reported that approximately 970 ships per day were experiencing GPS jamming in the Arabian Gulf and Strait of Hormuz across a four-day disruption window. The MSC Antonia grounding near Jeddah in May 2025 gave the problem a sharper operational edge: subsequent maritime intelligence analysis attributed the incident to GPS interference that distorted the vessel’s displayed or reported position. The dual function is the point. These operations probe navigational dependency and fallback capability, and they can manufacture legal or operational ambiguity without ever touching the vessel’s internal network. The selection logic is explicitly geopolitical: which vessels, on which routes, under which flags, are most exposed to navigational manipulation with the least recourse?
  5. Operator profile and shore-side maturity weigh in wherever adversaries hold any intelligence on the target organisation. Third-party managers running large, heterogeneous fleets across multiple flags are attractive precisely because security standardization is hard to enforce: a manager running 40 vessels across four ship managers, three VSAT providers, and two IBS vendors is a complex, unevenly hardened environment. A patient actor, nation-states have the patience, finds the weakest segment and enters there.
  6. Connectivity infrastructure is the most tactically decisive criterion, and it turns on a single question: does a shared provider serve the whole fleet? Where it does, provider compromise is operationally equivalent to compromising every vessel at once. The logic Lab Dookhtegan proved against Fanava Group, examined in detail below. Identifying a fleet’s shared provider costs an adversary almost nothing; the payoff is fleet-wide.
  7. Crew competency signals are an underappreciated dimension of target selection for actors pursuing initial access via the vessel itself rather than via shore-side infrastructure. Vessels operating with minimal dedicated IT personnel, which describes the majority of the global commercial fleet, are structurally reliant on crew members to perform tasks that in a shore environment would be handled by specialists. Chart updates on ECDIS via USB stick, remote diagnostic sessions authorized by an officer with no security training, crew welfare internet traffic running across the same VSAT link as operational systems: these are not hypothetical exposures. They are documented initial access vectors. Mustang Panda’s documented USB-based malware delivery aboard cargo vessels [confirmed by ESET analysis] is the operational expression of this insight.

Access Sequencing

Once a target is selected, the adversary works the access problem. The entry path is not chosen arbitrarily, it follows from the target’s exposure profile, the adversary’s objective, and the relative cost of each vector. Anticipating that logic is what lets a defender predict where pressure lands first.

1 – Shore-side compromise is the preferred path for patient, capability-rich actors, for a structural reason: ship managers, port agents, crewing agencies, classification societies, and technical managers all hold assets that convert directly into vessel access. Specifically, VPN credentials for remote diagnostic platforms, domain accounts reaching fleet management software, SFTP endpoints for chart and software distribution, and mailboxes full of voyage schedules and system configuration data. The adversary who compromises one of these inherits a relationship of implicit trust with the vessel. They are not an external entity probing a boundary, they are a known party whose access is operationally expected. SideWinder’s 2024–2025 campaign shows the reconnaissance depth this requires: lures built around port authority and regulatory correspondence [content demanding real knowledge of how maritime organizations communicate] delivered through CVE-2017-11882 [NVD – cve-2017-11882] in Office Equation Editor and a .NET downloader chain to StealerBot. The vulnerability is old, the targeting knowledge is current. That asymmetry is where these actors live.

For state actors with collection objectives, classification societies are a target in their own right, not a stepping stone. Class records hold construction drawings, system architecture, machinery specifications, and deficiency histories. A fleet-wide technical intelligence baseline that would take years to reconstruct ship by ship. Compromising the document management platform of a classification society can be the objective itself.

2 – VSAT as the initial access surface carries its own logic, evaluated separately from shore-side entry. Pre-exploitation reconnaissance is largely passive and Internet-facing: Shodan and Censys indexing of maritime SATCOM equipment [Cobham SAILOR, Intellian, KVH, Hughes, iDirect] routinely surfaces management interfaces, web UIs, Telnet, and SNMP endpoints on routable addresses, and the firmware versions they disclose can be cross-referenced against vulnerability databases without sending a packet to the target. Factory-default credential reuse is endemic, a meaningful proportion of internet-reachable terminals still answer to default authentication, exposing antenna control units, modem configuration, and routing tables.

The higher-leverage vector [the one Fanava demonstrated] is the provider hub, not the terminal. ST Engineering iDirect NMS (Network Management System) is a single administrative plane spanning an entire fleet: from it, an adversary can enumerate every terminal, push firmware and configuration changes, alter routing, observe or modify transiting traffic, and destroy the Falcon stack on every modem at once. The point defenders consistently miss is that in a hub-and-spoke topology, blast radius is not a property of the vessel’s posture at all. A perfectly hardened ship [current firmware, segmented networks, disciplined crew] is compromised the moment its provider’s NMS is, and nothing it does locally changes that. The terminal is a spoke; spokes inherit the security of the hub, and the hub is someone else’s network. IOActive, Inc. 2014 research framed maritime SATCOM exposure as a terminal problem. A decade later, Fanava reframed it as a management-plane problem, harder to defend and far more valuable to attack. The campaign still turned on outdated, poorly-credentialed iDirect terminals, mapped across months of patient access before the payload fired.

3 – Remote access exploitation is the most underweighted vector in maritime threat models and one of the most structurally exposed. Engine manufacturers, automation vendors, and IBS providers maintain persistent remote access to shipboard OT for diagnostics. Typically a VPN tunnel over the same VSAT link carrying crew welfare traffic, terminating on Windows engineering workstations that straddle the IT/OT boundary. Segmentation between those workstations and PLC-adjacent systems is inconsistent, and frequently absent. The credentials are held by the shore-side vendor and a shipboard point of contact. Usually a chief engineer or ETO with no security training and no way to know their credentials have been harvested. Compromise the vendor’s IT environment, and you inherit authenticated, trusted access to every vessel in its service base, invisibly to the operator. The 2023 DNV ShipManager ransomware incident [roughly 1,000 vessels affected through one platform] is the closest public marker of what this exposure produces at scale.

4 – AI-accelerated access sequencing has changed the economics of the access problem for well-resourced adversaries. The GTG-1002 campaign, disrupted by Anthropic in November 2025 after a September operation against roughly 30 global organizations. It was the first confirmed case of a Chinese state-sponsored actor running an autonomous AI framework through up to 90% of the intrusion lifecycle: reconnaissance, vulnerability identification, lateral movement, privilege escalation, credential harvesting, exfiltration, with humans intervening only at strategic decision points. It used open-source pentest tooling rather than custom malware, defeating signature-based detection, and compressed the gap from initial access to exfiltration to hours. The maritime read-across is direct: a fleet operator’s external surface exposed VSAT management interfaces, fleet portals, VPN endpoints, indexed maritime databases. It’s enumerable by exactly these techniques, in parallel, across an entire operator estate at once. The framework finds the weakest credential, the oldest firmware, the least-monitored access path, and sequences against the highest-probability target faster than a human analyst finishes one reconnaissance pass. Machine-speed access against human-speed detection is not a future concern. It is the current environment.

5 – Crew as the social engineering surface rewards maritime operational knowledge over generic phishing capability. The pretexts that work are operationally specific and time-sensitive by convention: ECDIS chart update packages, port pre-arrival notifications, charterer vetting questionnaires, Class survey scheduling, Flag State inspection notices and the time pressure is the mechanism: an officer who delays creates an operational problem. The ECDIS chart update vector is the cleanest. Updates arrive externally, load via USB or network transfer, and apply without meaningful validation in most workflows; a package that mimics a legitimate provider’s file structure and naming passes visual inspection and lands on the navigational core of the vessel. Mustang Panda’s USB-based delivery against European cargo operators is this logic at fleet scale, and APT41’s maritime-specific targeting [UK, Italian, Spanish, Turkish, Taiwanese, and Thai shipping and logistics firms] confirms the principle: adversaries build maritime-specific lures because they out-perform generic phishing against the same targets.


Persistence and Dwell

Maritime OT environments impose specific technical constraints on adversary persistence that differ materially from shore-side IT environments. Understanding how adversaries adapt to those constraints is prerequisite to building detection that can find them.

1 – Connectivity intermittency shapes C2 architecture. A vessel in open ocean runs on bandwidth-constrained, high-latency VSAT, with high-bandwidth windows only during port calls and [depending on route] stretches of total connectivity loss. An adversary holding a shipboard foothold cannot rely on a persistent synchronous C2 channel, which forces specific design choices: locally resident implants with scheduled or event-triggered beaconing rather than continuous polling, data queued for batch exfiltration during high-bandwidth windows, and persistence state that survives extended offline periods. APT41‘s ShadowPad fits this profile [modular, plugin-based, configurable beacon intervals, no requirement for a live channel to hold the foothold] and its modularity lets operators push capability during connectivity windows rather than landing a full-featured implant at initial access, keeping the early-dwell footprint small. VELVETSHELL, the other APT41 tool seen in maritime deployments, adds process-based detection evasion on the Windows endpoints common in IBS environments.

For implants that need periodic connectivity, port calls are the gift: shore cellular or WiFi supplements VSAT at predictable intervals that align with the vessel’s published AIS schedule. An adversary who has mapped a target’s port rotation [trivially available from AIS and commercial voyage-tracking data] times check-in and exfiltration to port arrival, blending outbound transfers into the elevated connectivity that already characterizes a port call.

2 – Passive enumeration is the adversary’s first move inside the operational network. Modbus TCP, DNP3, and NMEA 0183/2000 dominate vessel automation, and each is a reconnaissance surface. Modbus [still standard on ballast control, fuel management, and machinery monitoring] has no authentication: any device on the segment can issue function code 01 (Read Coils), 03 (Read Holding Registers), or 43 (Read Device Identification) against any reachable PLC and get an answer, no challenge. An adversary who has crossed from IT into the automation network [via a weak VLAN boundary or a dual-homed engineering workstation] can passively enumerate every Modbus device, map register ranges, and reconstruct which PLCs drive which physical processes, generating nothing a signature-based IDS would flag.

NMEA 0183 [serial-lineage, now routinely tunneled over TCP/UDP in modern IBS] has no authentication, no integrity checking, and no way to separate a real sensor reading from an injected one. GPS, AIS, gyrocompass, and echo sounder converge on the ECDIS through a multiplexer: a talker/listener fan-in where every device trusts every sentence on the bus. An adversary on that segment reads the whole navigational stream passively [position, heading, SOG, AIS contacts] and manipulation needs only well-formed sentences, a $GPGGA or $GPRMC with altered fix data and a correct checksum (a trivial XOR of the bytes, not a cryptographic control). The detail worth sitting with: on the wire, an injected sentence and a legitimate one are identical. No field encodes provenance. Detection cannot live at the protocol layer, because the protocol has no concept of authenticity. It has to live in cross-correlation: does the GPS position agree with radar, with the inertial system, with the expected track? That is precisely the layer most installations never instrument. The gap is not an oversight; it is the design assumption of a 1983 protocol built for a trusted serial loop, now carrying safety-critical data on a routable network. NMEA 2000 (CAN-bus derived), Modbus, and DNP3 inherit the same trust-the-bus premise, and retrofitting authenticity is largely impractical. Which is why mature maritime OT defense concentrates on segmentation and traffic-baseline anomaly detection. You are not going to fix NMEA. You are going to watch what speaks it.

Article content
In the bridge sensor chain, data flow becomes trust before authenticity is ever proven.

3 – The IT/OT boundary is thinner than the network diagram claims. The engineering workstations ETOs and chief engineers use for monitoring and remote diagnostics are Windows boxes bridging the IT VLAN and one or more OT VLANs, running fleet management clients, vendor diagnostic tools, and ECDIS chart management software side by side. A compromised one can yield domain credentials, shore-side VPN credentials, SFTP/FTP keys for chart and software distribution, local auth for PLC web interfaces and HMIs, and SSH keys for Linux automation systems. This is the documented access pattern in multiple confirmed maritime OT compromises, and it is why the engineering workstation is the priority lateral-movement target once an adversary holds the vessel IT network.

4 – Forensic visibility gaps are structural, not incidental. Most commercial vessels run without persistent traffic capture, centralized log aggregation, or SIEM. OT syslog, where configured at all, lands in local storage with short retention; Windows event logs on engineering workstations are not forwarded; VSAT traffic is not captured at the modem. An adversary doing passive OT recon, holding a scheduled-beacon implant, and exfiltrating during port calls leaves no persistent forensic record under normal operating conditions. And does not need to exploit that gap actively, only avoid the few behaviors that trip a local alert, which is easy when the alert surface is minimal. Read in this light, Lab Dookhtegan’s destruction of the Falcon stack and core system data in August 2025 was not forensic denial, the record was already thin. The wipe was operational: it blocked recovery and stretched the disruption by forcing full reinstallation on every vessel. Forensic denial was a side effect.

Fanava makes the sharpest version of the point. The five-month dwell between hub compromise and execution produced no vessel-level indicators, because the foothold never lived on a vessel. It lived in the provider’s management plane. Defenders watching ship systems for IOCs were watching the wrong environment. Management plane and operational plane are forensically distinct, and the adversary chose the one nobody was instrumenting.


Objective Execution

When adversaries move toward their objective, they time it against the vessel’s operational state and the broader geopolitical or operational context.

  • Data exfiltration objectives are typically executed during normal operations, when outbound data flows are expected and unlikely to trigger anomaly detection. Cargo manifests, voyage plans, crew records, commercial contracts, and technical documentation are high-value targets for intelligence collection. Mustang Panda’s campaigns against European cargo operators [Norway, Greece, the Netherlands] are consistent with this objective: the targets hold commercial intelligence on cargo flows, counterparty relationships, and trade routes that have strategic value to a state actor mapping global supply chains.
  • Disruptive objectives are timed for maximum operational and financial impact. Lab Dookhtegan’s March 2025 attack was explicitly timed to coincide with Operation Rough Rider, the US air campaign against Houthi forces in Yemen, compounding geopolitical pressure on Iran. The August 2025 wave landed amid intensifying US scrutiny and sanctions of Iran’s oil-export and shipping networks, context that amplified the impact of an operation that was already technically effective. Criminal ransomware actors follow a different timing logic: they typically execute payload deployment when the operator is most likely to pay quickly, which often means during port calls, cargo loading/unloading windows, or at the start of a voyage with a time-charter deadline.
  • Destructive objectives [the highest-stakes category] include physical manipulation of vessel systems: ballast valve control, propulsion interference, ECDIS chart data manipulation. Cytur’s 2026 maritime cyber reporting placed destroyed equipment, manipulated ECDIS chart data, and ballast-control interference inside the current vessel-risk discussion rather than the purely theoretical category. Public technical detail remains uneven, but the direction of travel is clear: adversary interest is moving closer to the operational systems that convert data trust into physical consequence. These are not conceptual scenarios. An adversary with access to a vessel’s OT network and the patience to understand the system architecture has the technical capability to create conditions that produce physical consequences. The timing logic for destructive execution is typically either maximum isolation (open ocean, minimal rescue capability) or maximum consequence (congested waterway, collision risk).

Defender Implications

The adversary target selection and access logic described above translates into specific defensive posture priorities.

1 – The supply chain is the threat boundary, not the vessel. The Fanava Group compromise is the clearest recent demonstration of this principle. If your VSAT provider, remote monitoring vendor, or technical management platform has access to your vessel systems, their security posture is part of your threat surface. Mapping third-party access paths to OT environments [what credentials exist, what systems they can reach, whether network segmentation exists between their access path and critical OT] is a prerequisite for understanding your actual exposure. Do this work before an adversary does it for you via Shodan.

Fleet defenders should demand security telemetry from providers, not only service availability metrics. For VSAT, remote diagnostics, and fleet management platforms, the relevant questions are not limited to uptime and SLA compliance. They are: who authenticated to the management plane, from where, against which terminals, with what privilege, and what configuration or firmware actions were performed? If the operator cannot receive or audit that telemetry, provider compromise remains invisible until operational impact reaches the vessel.

2 – Flag state and regulatory posture create detection asymmetry. Vessels flagged under weak inspection jurisdictions face lower detection probability for adversary access attempts, which makes them more attractive initial access targets within a mixed-flag fleet. If you operate a heterogeneous fleet, your threat hunting priority should account for the security posture variance across that fleet, not assume uniform hardening.

3 – OSINT reveals your exposure before you do. The combination of AIS data (vessel position, flag, destination, operator), Shodan enumeration (VSAT terminal make, model, firmware, exposed management interfaces), maritime databases (cargo type, charter party history, commercial relationships), and dark web marketplace data (voyage logs, cargo manifests, crew records already sold) gives a capable adversary substantial target intelligence without touching your systems. Run this reconnaissance against yourself. Understand what an adversary with no access can already see about your vessels and your fleet.

4 – Crew social engineering surfaces require operational context-aware training. Generic phishing awareness training does not address the specific pretexts adversaries use in maritime environments. ECDIS chart update requests, port authority clearance documentation, customs and charter correspondence. These require training that models maritime-specific attack scenarios, not generic enterprise phishing examples. If your crew security awareness program looks like it was built for a bank’s employees, it is not fit for purpose.

5 – Persistence in OT environments is quiet. An adversary who has compromised a vessel’s operational network and is conducting passive reconnaissance will not generate alerts. They will observe. Detection requires active threat hunting against the OT environment, looking for unexpected device enumeration, anomalous inter-system communication, lateral movement between IT and OT network segments, rather than relying on signature-based alerting. In most vessel environments, this hunting capability does not exist. Building it requires starting with network visibility: if you cannot see OT traffic, you cannot hunt through it.

6 – AI-accelerated adversary reconnaissance has collapsed the targeting timeline. The GTG-1002 model [autonomous vulnerability discovery, parallel target assessment, machine-speed lateral movement] is not confined to the sectors it first targeted. The same framework applies to any environment where adversaries can enumerate exposed services, test credentials, and identify OT network adjacency via internet-facing infrastructure. Maritime environments, with their high density of internet-exposed VSAT terminals, remote access platforms, and legacy OT systems, are a near-ideal application surface. The defensive implication is that exposure management [knowing what is visible and accessible from the internet, and closing or monitoring those surfaces] is now a time-critical function, not a periodic audit task. An adversary with AI-assisted reconnaissance capability can characterize your external attack surface faster than a quarterly review cycle can respond to it.

7 – Timing of execution is an intelligence problem, not just a detection problem. Adversaries time their execution against geopolitical events, operational windows, and financial pressure points. This means threat intelligence that tracks geopolitical developments affecting your trade routes and flag states is operationally relevant, not an abstract strategic concern. An operator whose vessels regularly transit Iranian-contested waters or Baltic chokepoints should have geopolitical risk integrated into their maritime security threat model, because the adversaries operating in those spaces demonstrably do.


Sources & further reading

Lab Dookhtegan / Fanava Group compromise

  • Cydome, “Lab Dookhtegan Cyberattack – Second Wave Findings (Aug 2025)” – cydome.io
  • Industrial Cyber, “Lab Dookhtegan cyberattack on Iranian oil tankers traced to supply chain compromise of Fanava’s infrastructure” (2 Sep 2025)
  • Iran International, “Hackers disrupt communications of dozens of Iranian oil and cargo ships” (Aug 2025)
  • The Maritime Executive, “Hackers Disable Iranian Merchant Shipping Communications” (24 Aug 2025)
  • CIMSEC, “The Unwitting Fleet” (2026)

GTG-1002 – AI-orchestrated campaign

  • Anthropic, “Disrupting the first reported AI-orchestrated cyber espionage campaign” (Nov 2025) – primary report
  • The Hacker News, “Chinese Hackers Use Anthropic’s AI to Launch Automated Cyber Espionage Campaign” (15 Nov 2025)
  • Horizon3.ai, “The First AI State-Sponsored Attack: What It Means for Defenders”

GPS spoofing – MSC Antonia and Strait of Hormuz

  • Inside GNSS, “MSC Antonia Grounding in the Red Sea Attributed to Suspected GNSS Spoofing” (15 May 2025)
  • gCaptain, “Pole Star Confirms GPS Interference Caused MSC Antonia Grounding” (15 May 2025)
  • The Maritime Executive, “Constant GPS Jamming Disrupts Navigation in Strait of Hormuz” (26 Jun 2025)
  • Scientific American, “GPS spoofing is scrambling ships in the Strait of Hormuz”
  • Windward, Q1 2025 maritime risk reporting (via above)

SideWinder APT

  • Kaspersky Securelist, “SideWinder APT attacks in H2 2024” (Mar 2025)
  • The Hacker News, “SideWinder APT Targets Maritime, Nuclear, and IT Sectors Across Asia, Middle East, and Africa” (Mar 2025)
  • BlackBerry Threat Research, maritime targeting analysis (Jul 2024)

Mustang Panda

  • ESET, “APT Activity Report Q4 2024–Q1 2025” – welivesecurity.com
  • The Maritime Executive / SAFETY4SEA, “China-linked cyber espionage group targets commercial shipping companies” (ESET, May 2024)

APT41, APT28, Crimson Sandstorm

  • SecurityWeek, “Chinese Hacking Group APT41 Infiltrates Global Shipping and Tech Sectors, Mandiant Warns” (Jul 2024)
  • Cyble, “Cyber Threats Surge Against Maritime Industry in 2025” (Jul 2025)

DNV ShipManager

  • SecurityWeek, “Ransomware Attack on DNV Ship Management Software Impacts 1,000 Vessels” (Jan 2023)
  • TechCrunch, “Maritime giant DNV says 1,000 ships affected by ransomware attack” (Jan 2023)

Maritime OT threat landscape – ECDIS, ballast, statistics

  • Cytur, “2026 Maritime Cyber Threat White Paper” (via The Maritime Executive and SAFETY4SEA, Feb 2026)
  • Industrial Cyber, “Cydome report finds 150% surge in maritime OT cyberattacks as ransomware tightens grip in 2025” (Mar 2026)

Foundational SATCOM security research

  • IOActive (Ruben Santamarta), “A Wake-up Call for SATCOM Security” (Black Hat, 2014)

Maritime Threat Craft is an independent LinkedIn article series written for maritime IT/OT security practitioners. Series identity: Adversary Thinking for Maritime Defenders. It is published by Alexandros Engelen, a Cybersecurity Strategist, specializing in Maritime Cyber Risk.


Revised: May 2026 – updated to incorporate GPS spoofing at operational scale (MSC Antonia, Persian Gulf 2025), GTG-1002 AI-autonomous access sequencing, and confirmed Fanava Group supply-chain forensic analysis.

The Invisible Occupation: Volt Typhoon and the Strategic Pre-Positioning of Maritime Infrastructure

Maritime Cyber Threat Briefing #11

Article content
From Quiet Access to Crisis-Time Disruption

In December 2024, something remarkable happened in a meeting room in Geneva. Senior Chinese officials reportedly signaled in a private Geneva meeting that the Volt Typhoon campaign was real, state-directed, and tied to Beijing’s military calculus around Taiwan. The Wall Street Journal

They did not use the name. They did not issue a statement. But the admission was clear enough that American officials left the room in no doubt: the campaign was real, it was state-directed, and it was linked to China’s military calculus around Taiwan.

That acknowledgement [which did not become public until April 2025] represents one of the most significant moments in the history of state-sponsored cyber operations. Not because China confessed. But because they felt compelled to say anything at all.

The Office of the Director of National Intelligence’s 2026 Annual Threat Assessment highlights persistent cyber threats posed by Chinese government-linked cyber actors. CISA, NSA, and FBI assess that these APT actors are positioning themselves within information technology networks, enabling lateral movement to operational technology systems [the hardware and software that control critical infrastructure] to disrupt critical functions at a time of their choosing. Control Risks

This briefing is about what that means for maritime. Because while the public discourse around Volt Typhoon has focused on power grids, water systems, and telecommunications, the maritime dimension of China’s pre-positioning strategy has received almost no attention in the shipping industry. It deserves urgent analysis.

The series has documented Iranian cyber-to-kinetic operations since Briefing #1, GPS spoofing and electronic warfare in Briefings #5 and #9, and port infrastructure attack methodology in Briefing #10. This briefing closes the arc on the geopolitical threat landscape by examining the most strategically sophisticated actor in the current threat environment and what it has already done, quietly, to maritime infrastructure that the global shipping industry depends on.


Understanding Volt Typhoon: This Is Not Espionage

Article content
Not Espionage. Pre-Positioning for Disruption.

The instinct, when confronting a Chinese state-sponsored cyber campaign, is to frame it as espionage, intelligence collection, data theft and commercial advantage. That framing is wrong for Volt Typhoon, and understanding why it is wrong is essential to understanding the maritime risk.

CISA‘s executive assistant director for cybersecurity explained that Volt Typhoon was found using living-off-the-land techniques on targets “where there is no reasonable espionage benefit.” When asked about the total number of victims, CISA’s executive director stated that any number given “is likely an underestimate.” Vectra AI

Volt Typhoon’s targeting and pattern of behavior have been assessed as pre-positioning to enable lateral movement to operational technology assets for potential destructive or disruptive attacks. The group has emphasized stealth using living-off-the-land binaries, hands-on keyboard activity, and stolen credentials. Cyble

The distinction is fundamental. An espionage campaign collects information that has value. Volt Typhoon is gaining access to systems whose value lies not in the data they hold but in the functions they perform:

  • Power distribution
  • Water treatment
  • Communications routing
  • Transportation management.

There is only one reason to pre-position inside operational technology systems that control physical infrastructure: to be able to disrupt or destroy those systems at a future moment of strategic choice.

In a secret meeting in Geneva in December 2024, Chinese officials indirectly acknowledged their role in cyberattacks against US critical infrastructure linked to the Volt Typhoon campaign. This admission validates long-standing fears that China’s cyber operations are designed to deter American intervention in Taiwan. Volt Typhoon represents a notable shift from cyber espionage to battlefield preparation, specifically targeting civilian infrastructure critical to US national resilience. Honeywell

The strategic logic is not difficult to decode. In a Taiwan crisis scenario, the United States would need to rapidly mobilise military assets across the Pacific. That mobilisation depends on functioning ports, functioning communications infrastructure, functioning logistics networks, and functioning fuel supply chains. Pre-positioning cyber access inside those systems [and being able to degrade or disable them at the moment of crisis] delays the American response, degrades military readiness, and potentially alters the outcome of a conflict without a single kinetic weapon being fired at a US military target.

This is not a theoretical scenario constructed by intelligence analysts. The scenario has been explicitly modelled: PRC forces move on Taiwan; the US Navy mobilises carrier groups; simultaneously, cyber sabotage cripples infrastructure on US territorial and allied islands, impacting power and water on key bases, delaying deployments, and jeopardising the Department of Defense’s ability to effectively aid Taiwan. Industrial Cyber

The maritime dimension of that model is direct, documented, and currently underweighted in the industry’s threat assessment.


The Maritime Target Set: What Has Been Confirmed

Article content
Maritime Infrastructure Is the Logistics Backbone of Crisis Response

Microsoft’s original disclosure explicitly listed the maritime sector in Volt Typhoon’s target set, and the US Coast Guard separately warned that the activity affected the Marine Transportation System. Microsoft United States Coast Guard

Microsoft’s original disclosure confirmed that Volt Typhoon has targeted organisations across communications, manufacturing, utility, transportation, construction, maritime, government, information technology, and education sectors. Volt Typhoon proxies all its network traffic to targets through compromised small office and home office network edge devices including routers from ASUS, Cisco, D-Link, NETGEAR, and Zyxel. PC Quest

Maritime is explicitly named in the confirmed target set. That is not incidental. The targeting of maritime infrastructure in Guam [a critical US military and logistics hub whose port facilities would be essential for any Pacific conflict mobilization] reflects a deliberate assessment of what needs to be degraded to slow the American response to a Taiwan crisis.

Beginning in 2021, Volt Typhoon executed a years-long cyber-espionage campaign that quietly targeted infrastructure on Guam, impacting telecom companies, satellite operators, and systems linked to US military operations on the island. The hackers exploited vulnerabilities in unsecured edge devices, directing attacks against Fortinet FortiGate appliances. Once inside, they harvested credentials and moved laterally across networks. The campaign remained undetected before being discovered by Microsoft researchers and US intelligence agencies. Smartmaritimenetwork

The Guam targeting is the clearest documented case of Volt Typhoon maritime-adjacent activity. But the implications extend far beyond Guam. US-allied ports in Japan, South Korea, Australia, and across the Indo-Pacific serve similar logistics and military support functions. The Kaohsiung and Keelung container ports in Taiwan [two of the world’s highest-throughput facilities] handle the semiconductor supply chain on which the entire global technology ecosystem depends. These ports have been assessed at a critical composite cyber risk score of 95 out of 100, facing threats from Volt Typhoon, APT41, and ArcaneDoor amid escalating China-Taiwan tensions. CongressGov

Beyond the immediate Taiwan theatre, Volt Typhoon’s confirmed activity across US West Coast transportation infrastructure, Hawaii, and Texas represents a systematic mapping of the maritime logistics network that would be required to sustain any extended Pacific military operation. These are not random targets. They are the nodes of a logistics chain, and the group has been inside them for years.


The Living-off-the-Land Methodology: Why You Cannot Find What You Cannot See

The most operationally significant characteristic of Volt Typhoon is not its targets. It is its technique and why that technique makes detection extraordinarily difficult in the maritime environment specifically.

Volt Typhoon abuses legitimate network administration tools used for routine maintenance, system configuration, and privilege management, making its activity indistinguishable from that of authorised administrators. As these actions rely on trusted tools and interfaces, traditional security controls often fail to detect anything unusual. This allows the threat actor to maintain long-term access while blending malicious traffic into normal network patterns. SecurityWeek

Living-off-the-land means exactly what it says. Rather than deploying custom malware that creates a detectable signature, Volt Typhoon uses the legitimate administrative tools already present in the target environment:

  • PowerShell
  • Windows Management Instrumentation
  • Command-line utilities

To conduct reconnaissance, harvest credentials, move laterally across networks, and establish persistence. From the perspective of any security monitoring tool looking for malicious software signatures, there is nothing to find. The activity looks like a network administrator doing their job.

When the FBI called the assistant general manager of Littleton Electric Light and Water Department on a Friday afternoon in November 2023, it was to tell him that Chinese state-sponsored hackers had been living inside his organisation’s systems for nearly ten months. The utility serves two small towns about 30 miles northwest of Boston. It is not a major defence contractor or a government agency. It is a local power and water company. Cyble

  1. Read that carefully. A local utility [not a defense contractor, not a government agency, not a strategically obvious target] had been hosting Volt Typhoon for ten months before the FBI found it and made the call. The utility had no idea. Their monitoring tools had nothing to flag. The attackers were using the utility’s own administrative tools to conduct their operations.
  2. Now apply that scenario to a maritime context. A port authority’s vessel traffic management system. A ship management company’s fleet operations platform. A satellite communications provider’s network operations centre. An OEM’s remote access infrastructure connecting to thousands of vessels. None of these are hardened government networks. Many are operated by organisations whose cybersecurity monitoring capability is limited. And the attacker’s methodology is specifically designed to be invisible to exactly the monitoring tools these organisations are most likely to have deployed.

CISA issued a supplementary advisory in February 2026 noting that Volt Typhoon activity had intensified since mid-2025, with new indicators of compromise identified in the water and communications sectors, characterising the heightened activity as consistent with “pre-conflict positioning.” CongressGov

The intensification in 2025, as US-China tensions elevated across the semiconductor export restriction dispute and the broader strategic competition, is not coincidental. The tempo of pre-positioning activity correlates with the geopolitical temperature. As the strategic environment deteriorates, the access is being expanded and verified.


The Taiwan Scenario: What Maritime Disruption Would Actually Look Like

The Taiwan Strait carries approximately 20 percent of global maritime trade. Kaohsiung is the fifteenth largest container port in the world. The semiconductor supply chain [on which every advanced military system, every data centre, every modern vehicle, and every consumer electronic device depends] flows through Taiwan’s ports.

The PLA has increasingly normalised military simulations of blockades and coercive maritime operations in the Taiwan Strait aimed at disrupting commercial traffic and asserting control over critical sea lanes. The CCP has ramped up cyberattacks against Taiwan’s energy grid. Taiwan’s National Security Bureau reported that successful cyber intrusions doubled year over year in 2024. Beijing is embedding itself in energy sector technology supply chains and targeting LNG terminal logistics to choke fuel offloading. Brside

Article content
Taiwan Crisis Activation Would Turn Hidden Access Into Maritime Disruption

Beijing might employ the China Coast Guard to undertake boarding, inspecting, and diversion operations under the guise of law enforcement. In combination with sabotage and cyberattacks on Taiwan’s critical infrastructure, as well as economic pressures on suppliers, Beijing might achieve an impact equivalent to a blockade without crossing the threshold of open armed conflict. Economy Middle East

The maritime cyber dimension of a Taiwan crisis would not be limited to Taiwan’s own ports and infrastructure. The pre-positioning already documented inside US transportation and communications infrastructure creates the capability to simultaneously degrade the response capacity of any nation attempting to support Taiwan, disrupting the port logistics, fuel distribution, and communications networks that military mobilization depends on.

A blockade would not be just an economic challenge. It represents a fundamental disruption to the global supply chain for advanced semiconductors, affecting manufacturers worldwide from German carmakers to US consumer electronics producers. The Taiwan Strait is a narrow but heavily traversed waterway where 20 percent of world shipping passes. Security Middle East

For commercial shipping companies, the practical consequence of a Taiwan Strait crisis [with or without the cyber dimension] is severe. Routes disrupted. War risk insurance premiums surging as they did in the Hormuz crisis. Cargo rerouted through longer and more expensive alternatives. Vessel scheduling thrown into months of uncertainty. The maritime industry learned this lesson from the Hormuz crisis in 2026. The Taiwan scenario carries the same economic consequences at potentially larger scale. The cyber pre-positioning documented inside maritime infrastructure adds a dimension of active operational disruption that the Hormuz crisis, for all its severity, did not fully demonstrate.


What Maritime Operators Need to Understand About Their Exposure

The Volt Typhoon threat to maritime infrastructure has three characteristics that make it distinct from every other threat category the series has previously documented.

  1. It is already inside. The Iranian cyber campaigns documented in Briefings #1 and #3 were active offensive operations, attacks that happened and were detected. Volt Typhoon’s methodology is designed for silent, persistent access that precedes any visible attack. The access is being established now, in environments that may show no signs of compromise, for use at a moment of Beijing’s choosing. The question for maritime operators is not whether this threat is coming. It is whether it is already present in their environment and they do not know it.
  2. The entry point is the edge device. Living-off-the-land attacks target edge devices such as routers, firewalls, and virtual private networks. From there, attackers access credentials of key personnel in networks without detection, enabling disruption of operational technologies. The 800 percent increase in edge device attacks documented in Cydome‘s 2026 Maritime Cyber Trends Report [highlighted in this series’ editorial research for Briefing #11] and Volt Typhoon’s documented methodology are describing the same attack surface from two different directions. The router sitting at the boundary of a vessel traffic management system, a port operations network, or a satellite communications hub is the entry point. And in maritime environments, these devices are frequently unpatched, undocumented, and unmonitored. Proarch
  3. The detection gap is structural. Traditional security monitoring looks for malicious software signatures. Volt Typhoon does not use malicious software. It uses the administrative tools the organization already has. Detecting it requires: behavioral monitoring, analyzing patterns of administrative activity and identifying anomalies against a baseline of normal behavior. CISA has confirmed that some Volt Typhoon victims are smaller organizations with limited cybersecurity capabilities that provide critical services to larger organizations or key geographic locations. The maritime supply chain is full of exactly these organizations: smaller port service companies, regional logistics providers, satellite communications resellers, and local vessel traffic management operators whose connectivity to larger maritime infrastructure makes them valuable entry points and whose security capability makes them accessible targets. Cydome

Mitigation: Hunting for What Signature-Based Tools Cannot Find

Article content
The Threat Hides Inside Normal Operations

The mitigation response to Volt Typhoon requires a fundamentally different security posture from the controls that address most other maritime cyber threats. Patching against known vulnerabilities matters. Fortinet FortiGate, Cisco IOS, Citrix ADC, and NETGEAR devices with unpatched CVEs are the documented initial access vector. But patching alone does not address an attacker that is already inside using legitimate tools.

  1. Threat hunting as an operational discipline. Volt Typhoon cannot be detected by waiting for alerts from signature-based tools. It requires active hunting. Deliberately searching for indicators of compromise and anomalous behavior patterns against a documented baseline of normal administrative activity. CISA advisory AA24-038A provides detailed indicators of compromise and YARA rules specifically developed for Volt Typhoon detection. Maritime organizations with critical infrastructure connections [port authorities, vessel traffic management operators, major shipping companies, communications providers] should be engaging threat hunting capability against these specific indicators as an operational priority, not a future aspiration.
  2. Edge device governance as a security critical function. Every router, firewall, VPN concentrator, and satellite terminal connected to a maritime operations network must be documented, inventoried, patched, and monitored. Management interfaces must not be exposed to the public internet. Default credentials must be eliminated. Volt Typhoon proxies all its network traffic through compromised SOHO network edge devices. Network edge devices owners should ensure that management interfaces are not exposed to the public internet to reduce the attack surface. This is a basic control. It is not universally implemented across the maritime sector, and the consequences of that gap are now documented at the highest levels of government intelligence. PC Quest
  3. Network segmentation between IT and OT as a containment measure. If Volt Typhoon gains access to the IT environment of a port or vessel management organization, the priority containment measure is ensuring it cannot reach OT systems. The vessel traffic management infrastructure, the terminal operating systems, the industrial control systems that govern physical operations. The lateral movement from IT to OT is Volt Typhoon’s documented next step after initial access. Segmentation does not prevent initial access. It limits the operational damage when access is achieved.
  4. Privileged access management and credential hygiene. Volt Typhoon’s primary technique after initial access is credential harvesting, collecting the administrative credentials that allow lateral movement across the network. Removing stale accounts, enforcing multi-factor authentication on all privileged access, limiting the scope of administrative credentials, and monitoring for unusual credential usage patterns are the controls that interrupt the lateral movement phase before it reaches OT systems.
  5. Intelligence-sharing participation. The Maritime Transportation System Information Sharing and Analysis Center in the United States, the Dutch Ferm Seaports collective intelligence model discussed in Briefing #10, and equivalent national maritime cybersecurity frameworks provide access to threat intelligence that individual maritime operators cannot generate independently. Volt Typhoon indicators of compromise, observed TTPs, and new advisory updates from CISA and Five Eyes partners are distributed through these channels. Maritime organisations that are not actively participating in these networks are operating without access to the most current available intelligence about the threat targeting their infrastructure.

Regulatory and Strategic Implications

Article content
Strategic Exposure Is Becoming a Maritime Governance Issue

The regulatory framework was not designed for a threat of this character. IMO MSC-FAL.1/Circ.3, IACS UR E26 and E27, and NIS2 were all developed with criminal ransomware operators and opportunistic state-sponsored espionage campaigns in mind. Volt Typhoon represents a different category: a patient, persistent, pre-conflict positioning campaign by a state actor that has explicitly acknowledged [in diplomatic back-channels if not public statements] that its purpose is to deter American military intervention in a Taiwan scenario.

That category requires a regulatory response that the maritime industry has not yet received. Current maritime cyber governance does not yet provide operational doctrine for confirmed state-sponsored pre-positioning inside maritime infrastructure. There is no classification society requirement for the threat hunting disciplines that Volt Typhoon detection requires. There is no mandatory reporting framework that would require a maritime organisation to disclose the discovery of Volt Typhoon access to flag state or port state control authorities.

NIS2 comes closest, its mandatory 24-hour reporting requirement for significant incidents and its supply chain security obligations provide a framework within which Volt Typhoon discovery would need to be reported in EU jurisdictions. But NIS2 does not reach the full scope of maritime infrastructure that Volt Typhoon is targeting, and its implementation remains uneven across member states.

The deeper strategic implication is one that the maritime industry needs to absorb as a collective: the shipping industry is not a neutral party in the geopolitical competition between major powers. It is logistics infrastructure. It is the network through which military mobilisation and economic pressure are both conducted. Pre-positioning inside maritime infrastructure is not an attack on a shipping company. It is an attack on the logistics capacity of the alliance that shipping company operates within.

That reframe has significant implications for how maritime organisations engage with national security agencies, how they approach intelligence sharing, and how they prioritise the security investments that protect not just their commercial operations but the broader strategic environment that makes those operations possible.


Article content

The Volt Typhoon campaign is one of the most strategically significant cyber threats ever documented against maritime infrastructure. It is not the most immediately visible, it is designed to be invisible. It is not the most operationally disruptive today, it is designed to be disruptive at a moment of Beijing’s choosing. It is not targeting data, it is targeting the ability to function.

By pre-positioning cyber threats within critical infrastructure networks, Volt Typhoon was poised to launch destructive cyberattacks of immense proportions. CISA confirmed that the group compromised critical infrastructure organizations in communications, energy, transportation systems, and water and wastewater systems. CongressGov

Transportation. Maritime is in that target set. It has been since 2021.

The maritime industry has built its cyber risk awareness around threats that announce themselves:

  • Ransomware that encrypts and demands
  • GPS spoofing that distorts navigation
  • Phishing that seeks immediate credentials
  • DDoS that overwhelms systems.

Volt Typhoon does none of these things. It waits. It maps. It holds access. And it is prepared to act at the moment when that access would have maximum strategic value. Which is precisely the moment when the maritime industry would be least able to absorb the disruption.

The Hormuz crisis gave the industry a live demonstration of what maritime hybrid warfare looks like when a determined state actor decides to weaponise a shipping chokepoint. The Taiwan scenario [with documented pre-positioned cyber access inside the maritime logistics infrastructure that would be essential to any response] raises that risk by an order of magnitude.

The question for every maritime organisation operating in the Indo-Pacific, connected to US-aligned logistics infrastructure, or serving as a node in the semiconductor supply chain, is not whether Volt Typhoon has them in its target set. The question is whether they are already inside and whether the organization would know.

Based on everything the intelligence tells us: many would not.


Maritime Cyber Threat Briefing is an independent series covering cyber threats, vulnerabilities, and risk management across the global maritime industry. It is published by Alexandros Engelen, a Cybersecurity Strategist, specializing in maritime cyber risk.

Sources: CISA Advisory AA24-038A, CISA China Threat Overview 2026, CISA ODNI 2026 Annual Threat Assessment, CISA Strengthening Resilience Against PRC Cyber Threats, Microsoft Security Blog [Volt Typhoon original disclosure], FBI Director Wray Congressional testimony January 2024, MITRE ATT&CK Group G1017, The Record / Recorded Future News [CISA RSA Roundtable], Industrial Cyber [IISS Volt Typhoon analysis], HSToday [Volt Typhoon critical infrastructure], Cybelangel [Volt Typhoon 2026 Active Report], Cyberwarzone [Volt Typhoon pre-positioning campaign], Georgetown SCS [Volt Typhoon infrastructure analysis], CYFIRMA [Taiwan cyber threat landscape], FDD [Chinese coercion Taiwan energy], Asia Times [China Taiwan quarantine strategy], NJCCIC Volt Typhoon analysis, Institute for Security and Technology [Cyber Redlines], Eclypsium [Chinese APT campaigns], CYDOME 2026 Maritime Cyber Trends Report, CYTUR 2026 Maritime Cyber Threat White Paper.

No commercial relationship exists with any cited organisation.