Act now, not later: understanding the legal risks of autonomous vessels before the MASS Code becomes mandatory

The world’s first international regulatory framework for crewless ships is being finalised in London this month. The IMO’s MASS Code, due for adoption at the Maritime Safety Committee’s 111th session as a non-binding instrument, sets a 10-year runway toward binding autonomous shipping standards. It is a landmark step. It is also, in important respects, the beginning rather than the end of the legal story.

Maritime autonomous surface ships (MASS) vessels, capable of operating with reduced or no crew, controlled remotely or by onboard systems, are no longer a distant prospect. Autonomous ferries are operating commercially in Scandinavia. Remotely operated vessels are transiting coastal routes in Asia. Trial programmes are underway in the UK, Japan, Singapore, and China. The technology is advancing faster than the law, and the gap between them is where liability risk accumulates.

What the MASS Code does and does not do

The code adopted at MSC 111 as a non-binding instrument establishes a goal-based framework setting out safety objectives for ships operating at varying degrees of autonomy, from vessels with automated decision support systems but a full crew, through to fully unmanned ships operating without any personnel on board. It is a structured starting point.

What follows is equally important. A three-year experience-building phase will gather operational data from vessels operating under that framework. The IMO will then begin drafting a mandatory MASS Code in 2028, with adoption targeted by July 2030 and entry into force expected on 1 January 2032. That is a six-year window during which autonomous vessels will operate commercially under a patchwork of national rules, flag state guidance, and conventions drafted for a world in which a master is always on the bridge.

The code addresses safety objectives. It does not resolve the harder legal questions, and it is those questions that will define whether autonomous shipping can be deployed with confidence.

The liability gap

The most significant unresolved issue is liability. Under existing maritime law, the framework for attributing fault in collision, grounding, or pollution cases assumes a human decision-maker. COLREGS, the International Regulations for Preventing Collisions at Sea, are built around the concept of a vessel under the command of a master exercising seamanship. The statutory duty to render assistance to persons in distress at sea presupposes a crew capable of doing so. When a vessel navigates by algorithm, the existing framework does not map cleanly onto the facts.

The MASS Code establishes functional requirements but leaves the question of liability allocation to flag states, coastal states, and ultimately courts. This is not a criticism of the IMO; liability was always going to be the hardest question to resolve by international consensus. But it means that the legal exposure for owners, operators, and insurers is real and present.

Take a straightforward scenario: an autonomous vessel in open sea collides with a fishing boat. There is no master to interview. No deck log written by a human hand. The onboard system made a navigational decision that turned out to be wrong. Who is liable: the owner? The operator of the remote control centre? The software developer whose algorithm determined the course alteration? The flag state that certified the vessel as fit for autonomous operation?

Existing cargo liability regimes create further complexity. The Hague-Visby Rules apportion responsibility by reference to the acts and omissions of the master and crew in the management and navigation of the ship. The navigational fault defence, one of the carriers’ most significant protections, requires a navigational error committed by a person. When navigation is automated, it is unclear whether that defence survives, or whether the carrier is left without it.

Regulatory fragmentation: the emerging risk

The delay between the non-binding code and the eventual mandatory framework creates fertile conditions for regulatory fragmentation. Flag states will interpret the code differently. Some will move quickly to establish domestic MASS frameworks that go further; others will apply existing rules with minimal adaptation. China, whose shipbuilding industry has a direct commercial interest in autonomous technology, finalised its own domestic regulatory framework for AI-managed vessel operations earlier this year, a move that has accelerated the IMO’s timeline, but also increases the risk that national standards diverge before international consensus is reached.

This fragmentation creates practical problems for internationally trading vessels. A vessel certified as autonomous under one flag state’s interpretation of the code may find itself in breach of port state control requirements in another jurisdiction. The genuine link between vessel and flag state, a foundational concept in UNCLOS, becomes complicated when the remote operations centre controlling the vessel is in a different jurisdiction entirely. Questions of jurisdiction, enforcement, and applicable law become considerably harder to answer.

For those financing autonomous vessel projects, these uncertainties matter. Lenders and lessors need to know what law governs their security, what events constitute a casualty for insurance purposes, and whether the regulatory framework in which the vessel operates will remain stable for the duration of the facility. At present, none of those questions have a definitive answer.

What operators and owners should be doing now

The absence of a mandatory framework does not mean the absence of legal risk. Parties operating in this space, or considering doing so, should be taking steps now.

• Review charterparty and bill of lading terms. Standard form contracts were not drafted with autonomous operation in mind. Terms around the master’s authority, safe port warranties, deviation, and the obligation to maintain a seaworthy vessel all require careful consideration. Parties should think through how existing clauses apply, or fail to apply, before a dispute arises rather than after.
• Examine insurance cover. P&I clubs and hull underwriters are beginning to engage with autonomous vessel risk, but cover is not uniform and the position is evolving rapidly. Owners and operators should not assume that existing policies respond to incidents involving autonomous systems in the same way they would to conventional casualties. Specific endorsements and exclusions need to be understood, and gaps identified before, rather than after, a loss.
• Embed cyber risk management. Port state control inspections now routinely check for cyber risk compliance within safety management systems, and underwriters are requiring evidence of cyber defence plans before covering autonomous or remotely operated vessels. This is not a future obligation; it applies now.
• Engage with flag state guidance. The code gives flag states discretion in how they implement its provisions. Owners considering operating autonomous vessels should engage early with their flag state to understand what certification pathway is available and what additional requirements may apply.
• Build a clear paper trail. In the absence of a human master, the records generated by autonomous systems, including logs, sensor data, and algorithmic decision records, will be the primary evidence in any dispute or casualty investigation. Ensuring that those records are preserved, accessible, and legally meaningful in any relevant jurisdiction is a practical step that should be built into operations from the outset.

A question of momentum

The MASS Code represents genuine progress. Its adoption marks the moment that the international community formally acknowledged that autonomous ships are coming and that the rules need to change. But the gap between the non-mandatory framework and the mandatory code, and the fragmented national landscape that will develop in between, means that legal certainty remains some years away.

For the industry, the 10-year runway is both an opportunity and a risk. Those who engage with the emerging framework proactively, who understand the legal exposure in the current grey area, and who structure their operations and contracts accordingly, will be better placed than those who wait for mandatory rules to arrive. The law will catch up with the technology. The question is what happens in the meantime.

For more information on autonomous vessel regulation or the issues discussed in this article, please contact our shipping team.