—
title: “The IMO MASS Code Is Here. Here Is What It Actually Says About Autonomous Ships, Liability, and the People Left Behind.”
date: “2026-05-22”
slug: “imo-mass-code-autonomous-ships-seafarers-liability-2026”
excerpt: “The IMO adopted the first international code for autonomous ships at MSC 111. This article breaks down what the code actually contains, what it leaves unresolved, and why every serving seafarer and maritime professional needs to understand it.”
tags: [“IMO MASS Code”, “autonomous ships”, “maritime regulation”, “seafarer welfare”, “MSC 111”, “MASS”, “maritime safety”, “human element”, “shipping technology”, “maritime law”]
image: mass-code-seafarers-hero.png
category: “Maritime Regulation”
—
The IMO MASS Code Is Here. Here Is What It Actually Says About Autonomous Ships, Liability, and the People Left Behind.
On 22 May 2026, the IMO Maritime Safety Committee adopted the world’s first International Code of Safety for Maritime Autonomous Surface Ships (MASS Code). After years of regulatory scoping exercises, working group sessions, and industry debate, there is now a formal international framework for autonomous commercial shipping.
The timing is striking. While the committee in London finalised rules for ships that may one day operate without human seafarers, an estimated 20,000 seafarers remain trapped aboard vessels inside the Persian Gulf. The contradiction between the push for unmanned ships and the industry’s inability to protect its current workforce is hard to ignore for anyone who has spent time at sea.
This article is not about whether autonomous ships are coming. They already exist. Fully electric autonomous container vessels are in commercial operation on short-sea routes in Norway. Coastal cargo ships in Japan, China, and Korea now use autonomous navigation and collision avoidance systems in daily service. Major shipbuilders are delivering newbuilds with autonomous ready systems integrated at the design stage. The technology is operational and scaling.
What this article examines is what the MASS Code actually contains, what it deliberately leaves unresolved, and what it means for the hundreds of thousands of seafarers who still crew the world’s merchant fleet.
What the MASS Code Actually Is and Is Not
The MASS Code adopted at MSC 111 is a goal based, non-mandatory regulatory framework. It sets safety objectives and functional requirements instead of prescribing detailed technical solutions. Flag states are not required to implement it, and shipowners face no immediate compliance obligations.
The Code is designed to supplement existing IMO instruments, particularly SOLAS. It provides a regulatory pathway for remotely controlled and autonomous ship operations while requiring that safety levels equivalent to conventional ships are maintained. It becomes available for voluntary application from 1 July 2026.
The adoption timeline is aggressive by IMO standards. An Experience Building Phase (EBP) will begin after MSC 112 in December 2026, during which real MASS operations will be tracked and assessed to build an evidence base. A mandatory MASS Code is targeted for adoption by 2030, with entry into force on 1 January 2032. The ECDIS mandate took over a decade from proposal to implementation. The MASS Code is moving much faster, driven by genuine commercial pressure and the risk of fragmented national standards filling the regulatory vacuum.
What the Code is not: It is not a liability framework, it does not resolve questions of criminal or civil responsibility when an autonomous vessel causes a casualty, and it does not address the workforce transition for conventional seafarers. These gaps are not oversight. They are deliberate boundaries set by the committee to keep the first version achievable.
The Four Degrees of Autonomy (and a Common Misunderstanding)
Underpinning the Code are four defined degrees of autonomy, carried forward from the IMO Regulatory Scoping Exercise:
- Degree 1: Ship with automated processes and decision support. Seafarers on board, ready to take control.
- Degree 2: Remotely controlled ship with seafarers on board.
- Degree 3: Remotely controlled ship without seafarers on board.
- Degree 4: Fully autonomous ship where the system makes decisions and takes actions without human intervention.
A significant clarification is needed here, and it has been causing confusion across the industry. Some operators, classification documents, and insurance questionnaires have begun asking vessels to declare a “degree of autonomy,” with the implication that any modern ship with integrated bridge systems and engine room automation qualifies as Degree 1. This is incorrect.
As West P&I has noted in recent guidance, a conventionally operated vessel with a full complement of crew and no remote control or autonomous navigation capability falls entirely outside the MASS framework. The appropriate declaration is “Not applicable.” Standard onboard automation assists the crew. It does not replace them, and it does not change the vessel’s operating model. The MASS degrees were created for the regulatory scoping exercise, not as a fleet wide classification system.
A vessel becomes a MASS only when automation formally changes the operating model by reducing, replacing, or relocating human control of key functions. That is a higher bar than merely having an integrated bridge system or an automatic engine room.
The 18 Chapter Framework
The MASS Code covers 18 areas, structured similarly to the Polar Code and IGF Code. Based on the structure finalised at MSC 111, the chapters address surveys and certification, risk assessment, system design, software principles, management of safe operations, alert management, connectivity and communications, remote operations centre requirements, structural integrity, fire protection and life saving, maritime security, search and rescue obligations, and environmental protection.
The Code takes a system of systems approach. Instead of certifying the entire ship as a single entity, it requires individual autonomous functions to be approved for specific operating envelopes: defined geographic areas, traffic conditions, weather limits, and cargo types. This means a vessel may be certified for autonomous operation in open ocean but required to revert to manned or remotely controlled operation in congested waters or restricted visibility.
Key provisions include requirements for:
- Connectivity and cybersecurity: Reliable, secure ship to shore communications with defined performance standards, latency limits, and backup modes. Protection against unauthorised access to control systems, with segregation between safety critical and non-critical networks.
- Fail safe states: Default to safe speed, stop, maintain heading, or follow a pre programmed safe pattern on loss of control or communications.
- Event data recording: Black box style logging of control commands, sensor inputs, and system states for casualty investigation.
- Sensor diversity: Multiple, diverse sensors including RADAR, AIS, GNSS, cameras, and LIDAR for situational awareness, with redundancy requirements.
- Operator intervention: Human operators must be able to intervene or take manual control where needed, within defined limits.
The Human Element: What the Code Addresses and What It Leaves Out
The human element chapter being finalised at MSC 111 addresses shore based remote operators: their certification, training, duty hours, and the interface between the remote control centre and the vessel. Under the Code, the remote operator in a Remote Operations Centre (ROC) can be functionally equivalent to a master, with the authority and accountability that implies. The Code requires clear assignment of roles and responsibilities between the ship and the ROC, documented procedures for handover of control, and fatigue management for remote watchkeepers equivalent to STCW requirements for seafarers on board.
This is necessary and overdue work. Remote operation of ships raises genuine human factors questions about situational awareness when the operator is looking at sensor feeds rather than out a bridge window, about cognitive load when monitoring multiple vessels simultaneously, and about the loss of the sensory inputs that experienced seafarers rely on, such as the feel of the ship, the sound of the engine, and the vibration of the hull.
But there is a significant gap the Code does not address: the hundreds of thousands of seafarers currently working aboard the conventional vessels that MASS adoption will partially replace over the coming decades. The transition risk is not mass immediate displacement. It is gradual attrition: fewer new sea going berths, a narrowing pipeline for the practical experience that underpins maritime safety culture, and an uncertain future for the career pathways that have drawn people to the sea for generations.
For the 2026 to 2032 window, Degrees 1 and 2 are expected to be the commercially dominant categories. These vessels will still require licensed officers, engineers, and ratings. But potentially fewer of them. The industry has already seen warning signs. The Q1 2026 Seafarers Happiness Index dropped to 7.01 from approximately 7.35, driven largely by the geopolitical pressures of the Persian Gulf crisis. Adding an uncertain regulatory transition on top of an already strained workforce is not a recipe for retention.
The Liability Gap the Code Does Not Close
This is perhaps the most consequential gap in the current framework. The MASS Code establishes functional safety requirements. It does not establish who is legally responsible when an autonomous vessel causes a collision, a grounding, or a pollution event.
The legal foundations of maritime liability were written for vessels with officers on the bridge making real time decisions. The collision regulations (COLREGS), the 1910 Brussels Convention on Collision Liability, and the existing liability conventions all assume a human master in command. When the person making the navigation decision is sitting in a shore based control centre in a different jurisdiction, and the systems executing that decision are driven by algorithms, the chain of accountability becomes unclear.
Is the remote operator the master? Is it the software developer? The flag state that certified the system? The company that deployed it? The Code flags these issues through goal based requirements but deliberately leaves the answers to flag state law, private law, and future IMO instruments.
Insurance markets are already flagging the problem. P&I clubs have begun noting that degree of autonomy declarations are appearing in documentation, and confusion over what they mean is creating potential for disputes. A shipowner who declares Degree 1 on an insurance questionnaire for a vessel that is simply a modern conventionally manned ship has created ambiguity about coverage. Conversely, a vessel operating genuine autonomous functions without proper declaration faces potential coverage gaps.
Until the liability framework catches up, operators carrying crews under Degrees 1 and 2 are operating in a legal grey zone that the MASS Code does not resolve. Every shipowner and operator trading under these conditions should be reviewing their P&I coverage and discussing the implications with their insurers.
Maintenance and Operational Reality
The MASS Code sets functional requirements for system design, software principles, and connectivity. It says comparatively little about how those systems are maintained when the ship is at sea and the crew has been reduced.
This matters because shipboard maintenance does not fundamentally change with automation. Sensors need cleaning. Algorithms need updating. Communication links need troubleshooting. Fire detection systems need testing. The hull and machinery still degrade whether there are ten people on board or two. In some respects, the complexity increases. An autonomous system with sensors, processors, actuators, and redundant communication links introduces more failure points than a conventional ship, not fewer.
If crew numbers are reduced, the remaining crew carry a higher maintenance burden. Fewer hands for the same physical work, plus new tasks that require different technical skills. The Code does not address whether the manning scales assumed in the STCW Convention and minimum safe manning requirements are adequate for vessels operating under partial autonomy. This is a question that port state control authorities and flag states will need to grapple with.
There is also the question of resilience in contested environments. The GPS jamming and spoofing documented in the Strait of Hormuz, the Black Sea, and the Eastern Mediterranean are not theoretical risks for autonomous vessels. They are existential ones. A conventional ship can fall back on a deck officer’s eyes, radar interpretation, and judgment developed over years at sea. An autonomous vessel running on algorithms and remote connectivity has no equivalent fallback if its primary positioning and communication systems are compromised. The Code includes cybersecurity requirements, but the threat environment evolves faster than the amendment cycle for a non-mandatory instrument.
On a related note, the Code requires even fully unmanned vessels to carry a plan for conducting search and rescue operations. An autonomous ship is not relieved of its SOLAS obligation to render assistance. The practical question the Code does not answer is how a vessel without crew puts a rescue boat in the water. This is not a minor detail. It touches on the fundamental legal and moral obligation of all ships to assist persons in distress, an obligation that has defined maritime culture for centuries.
What the Experience Building Phase Will Test
The voluntary period from July 2026 through approximately 2028 is not a pause. It is designed as an active data collection exercise. Flag states that choose to authorise MASS operations during this phase will be expected to report incident data, near misses, and operational performance back to the IMO. The results will directly shape the mandatory Code.
Several critical questions are expected to be tested during this phase:
- How do remote operators maintain situational awareness equivalent to a bridge watchkeeper over extended periods?
- What is the minimum crew required for a Degree 2 vessel, and can that be proven through operational data rather than theoretical models?
- How do autonomous systems interact with conventional traffic in practice, particularly in congested waters and restricted visibility?
- Can remote operations centres be certified and audited under something equivalent to the ISM Code framework?
- What happens to liability when an insured vessel operating under autonomy suffers a casualty?
The China Factor
No analysis of the MASS Code is complete without acknowledging the regulatory context in which it was adopted. China’s Smart Shipping Action Plan targets over 100 smart vessels by 2027, with 60 automated ports and a nationwide autonomous shipping ecosystem under development. The “Zhifei” autonomous container ship is already in service along the Chinese coast. Beijing has made it clear that if the IMO moves too slowly on mandatory regulation, China will codify its own domestic standards.
This is the real risk of regulatory fragmentation. The MASS Code’s voluntary nature, combined with the aggressive unilateral programs in China, Korea, Japan, and Norway, creates conditions where multiple national standards could emerge before the mandatory IMO framework arrives in 2032. For shipowners operating across multiple jurisdictions, the compliance burden could become significant.
What This Means for the Industry Now
The MASS Code is neither a threat to every seafarer’s job tomorrow nor a harmless paperwork exercise. It is the beginning of a regulatory transition that will unfold over the next decade. For shipowners, the immediate action is to understand whether any of their vessels fall within the MASS framework, ensure that insurance declarations are accurate and consistent with the actual operating model, and participate in the experience building phase where possible.
For seafarers, the message is more nuanced. The transition to autonomous shipping does not eliminate the need for maritime judgment. It concentrates it. The seafarer’s role shifts toward monitoring, troubleshooting, and systems management. The question is whether the training infrastructure, career pathways, and regulatory protections will be ready when the jobs begin to change.
The maritime industry has been through technological transitions before. Engines replaced sails. GPS replaced celestial navigation. Integrated bridge systems transformed watchkeeping. Each time, automation changed the nature of seafaring work but did not eliminate it. The MASS Code is the next chapter in that story. The outcome depends on whether the people with the most relevant experience are included in writing it.
The MASS Code was adopted at IMO MSC 111 on 22 May 2026. It becomes available for voluntary application from 1 July 2026. The Experience Building Phase begins after MSC 112 in December 2026. The mandatory Code is expected by 2030, with entry into force on 1 January 2032.
Written by Nittin Handa. Master Mariner, maritime policy professional based in Hong Kong.
This article reflects the author’s personal views and professional experience. It does not represent the position of any organisation with which he is affiliated.