1. Introduction

Autonomous vessel governance concerns the legal rules governing ships that navigate, monitor their surroundings, avoid collisions, manage propulsion, or perform other operational functions with limited or no direct human intervention.

These vessels are commonly described as Maritime Autonomous Surface Ships (MASS). They may be remotely controlled from shore, partially automated with a crew onboard, or capable of making operational decisions autonomously.

Civil law is relevant because autonomous shipping can generate disputes involving:

Collision damage, personal injury, and wrongful death.

Cargo loss, delay, and contractual performance.

Negligence and defective navigation software.

Shipowner, operator, manufacturer, and remote-controller liability.

Cybersecurity breaches and unlawful interference.

Marine pollution and environmental damage.

Insurance coverage and indemnity.

Port access, licensing, and regulatory compliance.

A defining feature of autonomous shipping is that a vessel may make a navigational decision without a human physically onboard. Nevertheless, autonomy does not automatically remove the legal responsibilities of the shipowner, operator, or other responsible parties.

Current international framework: In May 2026, the International Maritime Organization (IMO) adopted a non-mandatory International Code of Safety for Maritime Autonomous Surface Ships. The Code took effect on 1 July 2026 and provides a global framework for safety, risk assessment, remote operations, and human oversight. Work toward a future mandatory code remains underway.

International Maritime Organization

+1

An important qualification for case-law research is that courts have not yet developed a large body of decisions directly concerning collisions caused by fully autonomous commercial vessels. The analysis below therefore distinguishes established maritime judgments from broader precedents that are relevant by analogy.

2. Meaning and classifications of autonomous vessels

Level 1 — Shipboard assistance

Software supports navigation, route planning, collision alerts, and engine monitoring. The master and crew retain direct operational control.

Legal focus: Ordinary navigational duties, professional negligence, and equipment reliability.

Level 2 — Partial autonomy

The system executes certain manoeuvres or operational tasks while human personnel supervise and intervene when necessary.

Legal focus: Allocation of responsibility between the master, operator, and automated control system.

Level 3 — Remote-controlled vessel

Operators located ashore monitor and control the vessel through communications systems.

Legal focus: Remote master's authority, communications failures, fatigue, cybersecurity, and shore-based negligence.

Level 4 — Highly or fully autonomous vessel

The vessel performs substantial or all navigational functions without continuous human control, within its approved operational limits.

Legal focus: System design, algorithmic decision-making, foreseeable failure, product liability, and the attribution of fault.

These are descriptive categories rather than a substitute for the applicable IMO classification and certification requirements. A vessel's legal status depends on its actual functions, operating conditions, and regulatory approvals.

3. Principal international legal framework

A. United Nations Convention on the Law of the Sea (UNCLOS), 1982

UNCLOS provides the broad international framework for maritime jurisdiction, navigation, and State responsibilities.

Article 94: Flag States must effectively exercise jurisdiction and control over ships flying their flags, including relevant safety and operational requirements.

Article 98: States must require masters, so far as they can do so without serious danger to their vessels, crews, or passengers, to render assistance to persons in danger at sea.

Article 192: States have the general obligation to protect and preserve the marine environment.

Article 194: States must take measures to prevent, reduce, and control marine pollution, subject to the Convention's detailed provisions.

Autonomous vessels raise interpretive questions concerning how duties traditionally assigned to a master or crew should be performed when navigation is remote or automated. The use of autonomy does not itself eliminate the relevant international obligations.

B. International Regulations for Preventing Collisions at Sea (COLREGs), 1972

COLREGs are central to collision liability.

Important provisions include:

Rule 2: Responsibility and the need to exercise proper seamanship and precautions.

Rule 5: Keeping a proper lookout.

Rule 6: Proceeding at a safe speed.

Rule 7: Determining whether risk of collision exists.

Rule 8: Taking appropriate action to avoid collision.

Rules 13–17: Overtaking, head-on situations, crossing situations, and the responsibilities of vessels in particular encounters.

For autonomous vessels, the difficult question is how a system can satisfy standards such as keeping a proper lookout and taking timely collision-avoidance action. Relevant evidence may include sensor performance, software behaviour, the approved operational design domain, and whether a remote operator could intervene in time.

An autonomous vessel does not obtain an exemption from collision-prevention requirements merely because its decisions are made by software.

C. SOLAS Convention, 1974

The International Convention for the Safety of Life at Sea establishes major international standards for ship construction, equipment, operation, and safety.

Autonomous vessels raise questions concerning the functions of the master, onboard personnel, remote-control centres, emergency response, and communications reliability.

The 2026 non-mandatory MASS Code supplements the existing international safety framework and addresses gaps identified in applying conventional ship regulations to autonomous and remotely operated vessels. The IMO has outlined further work toward a mandatory code, including a target of adoption by 2030 and entry into force in 2032.

International Maritime Organization

+1

D. International Convention on Civil Liability for Oil Pollution Damage, 1992

Where a qualifying oil pollution incident involves a covered ship, this Convention may impose liability on the shipowner, subject to its scope, conditions, and exceptions.

Autonomous navigation does not automatically displace the applicable pollution-liability regime. The legal inquiry includes whether the vessel and damage fall within the Convention and whether the relevant conditions for liability and limitation are satisfied.

Other maritime pollution instruments, including MARPOL, may impose additional obligations depending on the incident.

E. Maritime contracts and domestic civil law

International conventions do not resolve every dispute between private parties. National laws and contractual arrangements may govern:

Charterparty and carriage-of-goods disputes.

Ship-management and remote-operation agreements.

Software licensing and maintenance contracts.

Negligent navigation and personal injury claims.

Product defects and cybersecurity failures.

Insurance, indemnity, and contribution between responsible parties.

The applicable law and forum depend on jurisdictional rules, contractual terms, statutory provisions, and the facts of the incident.

4. Comparative case law: eight important judicial authorities

The following decisions establish principles relevant to autonomous vessel governance. They concern conventional maritime navigation, seaworthiness, collision liability, and negligence—not direct judicial findings about fully autonomous vessels.

Case 1. The Pennsylvania (1873)

Court: Supreme Court of the United States.

Citation: 86 U.S. (19 Wall.) 125.

Facts: The case arose from a maritime collision in which a vessel's violation of a statutory navigation requirement was relevant to determining liability.

Legal principle: Under what became known as the Pennsylvania Rule, a vessel that violates a statutory rule intended to prevent collisions bears a demanding burden of showing that its violation could not have contributed to the collision.

Application to autonomous vessels: If an autonomous ship violates a collision-prevention requirement because its navigation software fails to identify a crossing vessel, that violation may be significant in subsequent litigation. The precise consequences depend on the applicable law and proof of the relevant facts.

This case is particularly relevant to the legal evaluation of algorithmic compliance with COLREGs.

Case 2. The T.J. Hooper (1932)

Court: United States Court of Appeals for the Second Circuit.

Citation: 60 F.2d 737 (2d Cir. 1932).

Facts: Tugboats were lost in a storm after failing to receive weather warnings that could have helped them avoid the danger. The case examined whether the failure to carry radio receiving equipment constituted negligence.

Legal principle: Industry custom does not conclusively determine the legal standard of reasonable care. A practice may be customary yet still be inadequate.

Application to autonomous vessels: A shipowner cannot necessarily defend an inadequate autonomous navigation or collision-avoidance system merely by arguing that similar systems are commonly used in the industry.

Courts may examine whether reasonable safety precautions were required in the circumstances, including sensor redundancy, communications reliability, testing, and emergency intervention capability.

Case 3. United States v. Reliable Transfer Co. (1975)

Court: Supreme Court of the United States.

Citation: 421 U.S. 397.

Facts: The case concerned a vessel grounding and the allocation of liability for maritime damage.

Legal principle: The Court replaced the traditional equal division of damages between mutually at-fault vessels with a comparative-fault approach, allocating damages according to the parties' respective degrees of responsibility where practicable.

Application to autonomous vessels: Where an autonomous vessel and a conventional vessel contribute to a collision, liability may depend on the relative fault of the responsible parties under the applicable maritime law.

The analysis may include negligent remote operation, deficient system design, inadequate maintenance, or failure to comply with collision-prevention rules. Allocation of responsibility to manufacturers or software providers may require separate legal analysis.

Case 4. The Wagon Mound (No. 1) (1961)

Court: Privy Council, on appeal from Australia.

Citation: Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co Ltd [1961] AC 388.

Facts: Oil spilled from a vessel into Sydney Harbour and later caught fire, damaging a nearby wharf. The case examined whether the type of damage was reasonably foreseeable.

Legal principle: The decision is a leading authority on remoteness of damage in negligence. The foreseeability of the relevant kind of harm is central to the analysis.

Application to autonomous vessels: If a software defect causes a collision that subsequently produces a fire or pollution incident, the court may need to determine whether the relevant damage was sufficiently foreseeable under the applicable law.

The case helps structure the causation inquiry but does not establish a special rule for algorithmic systems.

Case 5. Donoghue v. Stevenson (1932)

Court: House of Lords, United Kingdom.

Citation: [1932] AC 562.

Facts: The claimant alleged illness after consuming ginger beer that contained a decomposed snail. The dispute concerned whether a manufacturer owed a duty of care to the ultimate consumer.

Legal principle: The decision established the modern neighbour principle in negligence, recognising duties toward persons foreseeably affected by one's conduct, subject to the applicable legal requirements.

Application to autonomous vessels: The principle may inform analysis of whether manufacturers, software developers, and system integrators owe duties of care to persons foreseeably exposed to risks from defective autonomous navigation systems.

It does not mean that every software developer automatically owes every injured person a duty. The applicable jurisdiction's rules on duty, causation, and economic loss remain important.

Case 6. Scindia Steam Navigation Co Ltd v. De Los Santos (1981)

Court: Supreme Court of the United States.

Citation: 451 U.S. 156.

Facts: The case concerned a longshoreman's injury during cargo operations and the allocation of duties among a vessel owner and stevedores under the applicable maritime framework.

Legal principle: The decision examined the scope of a vessel owner's duties in relation to hazardous conditions and the responsibilities of other parties involved in cargo operations.

Application to autonomous vessels: An autonomous vessel does not operate in isolation. Responsibility for a cargo-handling injury may involve the vessel owner, terminal operator, cargo contractor, remote-control provider, and equipment manufacturer.

The case illustrates the importance of distinguishing each party's role, knowledge, contractual responsibilities, and legal duties rather than attributing every accident to the ship's automation.

Case 7. The Ocean Victory (2017)

Court: Supreme Court of the United Kingdom.

Citation: Gard Marine & Energy Ltd v China National Chartering Co Ltd [2017] UKSC 35.

Facts: The dispute arose from the loss of a vessel at a Japanese port in severe weather. It involved the allocation of risk between a shipowner, charterer, and insurers, including issues relating to the charterparty and the safe-port warranty.

Legal principle: The judgment examined contractual risk allocation and the application of safe-port principles in the circumstances of the casualty.

Application to autonomous vessels: Autonomous shipping contracts should identify operational limitations, acceptable environmental conditions, port-access requirements, and responsibility for decisions to enter or leave a port during dangerous weather.

The case is relevant to contractual allocation of maritime risk, not a direct ruling on autonomous navigation.

Case 8. Alize 1954 and CMA CGM SA v Allianz Elementar Versicherungs AG — The CMA CGM Libra (2021)

Court: Supreme Court of the United Kingdom.

Citation: [2021] UKSC 51.

Facts: The container ship CMA CGM Libra grounded after departing from Xiamen, China. The litigation concerned whether a defective passage plan rendered the vessel unseaworthy and the consequences for cargo interests.

Legal principle: The Supreme Court upheld the conclusion that the defective passage plan rendered the vessel unseaworthy for the purposes of the relevant carriage-of-goods dispute. A vessel's safe navigation involves proper preparation and planning, not simply the physical condition of its hull and machinery.

Application to autonomous vessels: A navigation system may be technologically sophisticated and still be unsafe if it relies on defective charts, inadequate route planning, incorrect environmental data, or an insufficiently tested operational model.

For autonomous vessels, seaworthiness assessments may therefore need to consider software, sensor integration, data quality, route planning, and the suitability of the system for the intended voyage.

This is one of the most useful modern maritime precedents for examining the relationship between navigational planning and vessel safety.

5. Comparative national approaches

Different jurisdictions apply established maritime principles through different statutory and regulatory frameworks.

This is a high-level comparison. The applicable law may differ according to the flag State, location of the incident, vessel type, nature of the claim, and relevant international conventions.

A. United Kingdom

British maritime law provides important guidance on seaworthiness, navigation, charterparties, and negligence. Decisions such as The CMA CGM Libra show that a vessel's safe operation depends on adequate navigational planning.

For autonomous vessels, the inquiry may extend to whether the navigation software was suitable for the voyage, whether route data were reliable, and whether the operator established appropriate safeguards.

B. United States

US maritime law offers established principles for collision liability and the allocation of damages. The Pennsylvania, The T.J. Hooper, and United States v. Reliable Transfer are particularly relevant.

A collision involving an autonomous vessel may require courts to determine whether the system violated a statutory navigation rule, whether the operator acted reasonably, and whether defective equipment or software contributed to the accident.

C. European Union

EU maritime governance combines international maritime conventions with applicable regional and national rules. Autonomous shipping also raises questions about data processing, cybersecurity, product safety, and cross-border coordination.

The legal challenge is to maintain equivalent safety standards while adapting operational requirements to vessels that may not carry conventional onboard crews.

D. Australia

Australian maritime law combines international obligations, shipping legislation, and negligence principles. The Wagon Mound is particularly important to the analysis of foreseeable damage, although its authority must be understood in its common-law context.

An autonomous vessel's operator may need to consider foreseeable collision, pollution, and property-damage risks when deciding whether to deploy a particular system.

E. India

India's maritime framework includes shipping legislation, applicable international conventions, contract law, and general civil liability principles.

A dispute involving an autonomous vessel may raise questions about registration, safe navigation, insurance, cargo damage, marine pollution, and the responsibility of the operator or technology provider. The outcome will depend on the governing statutory provisions and the facts of the incident.

6. Civil liability in autonomous vessel accidents

Responsibility may be distributed among several parties.

Shipowner

Responsible for applicable ownership, maintenance, seaworthiness, insurance, and operational obligations. Whether the owner is liable for a particular accident depends on the relevant legal regime.

Ship operator or remote master

May be responsible for unsafe operating decisions, inadequate supervision, failure to respond to warnings, or non-compliance with navigation rules.

Manufacturer or software developer

May face product-liability or negligence claims where a legally actionable defect or failure in design, manufacture, warnings, or software performance causes damage.

Communications and sensor providers

May be liable where defective services or equipment breach an applicable duty or contractual obligation and cause a compensable loss.

Flag State and regulators

Have relevant public-law and international obligations, but their responsibility is distinct from the private civil liability of commercial operators. State liability is not automatically established by every vessel accident.

The fact that a system made an autonomous decision does not itself determine which party is liable. A court may need to consider the design of the system, its operating conditions, human supervision, maintenance, and the causal relationship between the alleged failure and the damage.

7. Major legal challenges

A. The meaning of the master

Traditional maritime rules frequently assign duties to the master. Autonomous vessels challenge the assumption that the person exercising navigational authority must be physically onboard.

A workable governance system must identify who has command authority, who can intervene, and who is responsible when the vessel enters an emergency situation.

B. Compliance with collision-prevention rules

COLREGs rely on concepts such as proper lookout, safe speed, and action based on good seamanship. Autonomous systems must implement these duties reliably in complex traffic conditions.

The relevant evidence may include sensor records, decision logs, software versions, and simulations of the encounter. Compliance cannot be assessed merely by showing that an algorithm was installed.

C. Seaworthiness and software defects

The traditional concept of seaworthiness extends beyond the physical condition of the vessel. Autonomous navigation introduces questions about software validation, data quality, cybersecurity, and the adequacy of the overall operational system.

The reasoning in The CMA CGM Libra is relevant because it highlights the legal importance of proper navigation planning.

D. Causation and proof

An autonomous vessel may produce large quantities of technical data, but that does not necessarily make the cause of an accident easy to establish.

Investigators may need to determine whether the incident resulted from a software defect, poor sensor performance, inadequate maintenance, a communications failure, operator error, or a combination of factors.

E. Limitation of liability and insurance

Maritime law may permit limitation of liability in qualifying circumstances. Whether limitation is available depends on the relevant convention or domestic statute and the applicable conditions.

Insurance contracts should clearly address autonomous navigation, remote-control failures, cyber incidents, equipment defects, and allocation of liability among owners, operators, and technology suppliers.

8. Practical governance safeguards

A legally robust autonomous vessel framework should include:

Risk-based certification: Demonstrate that the vessel is safe within its intended operating conditions.

Human oversight: Identify the responsible master or operator and define intervention procedures.

Collision-avoidance validation: Test compliance with applicable navigation rules under realistic conditions.

Cybersecurity controls: Protect navigation systems, communications, and command authentication.

Event recording: Preserve navigational decisions, sensor data, software changes, and operator interventions.

Clear contracts: Allocate responsibilities among shipowners, operators, manufacturers, remote-control providers, and insurers.

Emergency procedures: Establish responses to communications loss, sensor failure, unsafe weather, and unexpected obstacles.

Environmental safeguards: Incorporate pollution prevention, waste management, and appropriate end-of-life procedures.

Effective remedies: Preserve appropriate routes for injured persons, cargo interests, and other affected parties to pursue claims.

9. Conclusion

Civil law and comparative autonomous vessel governance concern the adaptation of established maritime principles to ships capable of navigating and operating with reduced human involvement.

The central conclusions are:

Autonomy does not eliminate legal responsibility. Existing rules on navigation, seaworthiness, negligence, contracts, and maritime liability remain important.

Liability depends on the applicable law and evidence. The roles of shipowners, remote operators, manufacturers, and other parties must be assessed rather than assuming that the algorithm alone is responsible.

International standards and domestic remedies must work together. UNCLOS, COLREGs, SOLAS, pollution conventions, national legislation, and maritime contracts address different parts of the legal problem.

Governance must be proactive. Certification, system testing, human oversight, cybersecurity, audit records, and insurance are essential to managing risks before an accident occurs.

The eight judicial authorities discussed—The Pennsylvania, The T.J. Hooper, United States v. Reliable Transfer, The Wagon Mound, Donoghue v. Stevenson, Scindia Steam Navigation, The Ocean Victory, and The CMA CGM Libra—provide a substantial comparative foundation for analysing autonomous vessel governance.

They do not collectively constitute eight precedents directly deciding autonomous-ship liability. Rather, they establish maritime and civil-law principles that courts and regulators can apply, adapt, or distinguish as autonomous shipping develops.