Splash247: Shipping’s AI race runs ahead of the rulebook

Published by Splash247

Shipping is adopting AI faster than the rules meant to govern it, though the regulatory picture is less empty than it was even a year ago.

The clearest marker came on 22 May, when the International Maritime Organization’s Maritime Safety Committee adopted the International Code of Safety for Maritime Autonomous Surface Ships, the first global framework covering remote-controlled and AI-enabled cargo vessels. The MASS Code entered into force on 1 July as a voluntary instrument, opening what the IMO calls an “Experience Building Phase” during which flag states, owners and class societies gather data before any mandatory rules follow. IMO secretary-general Arsenio Dominguez said the code “positions IMO at the forefront of regulating emerging technologies.”

The catch is scope. The MASS Code governs degrees of ship autonomy, from decision-support systems with crew aboard through to fully autonomous vessels. Regulators define four such degrees: automated systems with seafarers on board to operate them, remotely controlled ships with crew available to intervene, remotely controlled ships with no crew at all, and fully autonomous vessels capable of making decisions independently. Remote operations centres themselves fall under the code too, subject to flag-state approval. But the framework says little about the AI tools already spreading fastest through the industry: onboard assistants, predictive maintenance systems and the administrative software captains and shore staff use daily. That layer remains governed unevenly, company by company and class society by class society, rather than by a single binding standard.

Classification societies are filling some of the gap themselves. DNV last month launched RuleAgent, an AI tool that searches its own rules database and links every answer back to the source text rather than generating unverified responses. Cristina Saenz de Santa Maria, DNV’s interim maritime chief executive, said the business was “committed to leading this development in a responsible and practical way.” Project manager Martin Borge Heir said the goal was “to provide a trusted reference for rule-related questions,” with results traceable to official sources rather than taken on faith. DNV describes its AI tools as assistive by design, built to be checked rather than followed blindly.

That instinct matches what shipping companies are telling their own suppliers. A recent survey of 300 seafarers and officers by maritime IT provider Dualog found little resistance to AI itself, with chief commercial officer Silje Moan saying seafarers were already experimenting with the technology and wanted help solving problems. The constraint, she said, was trust rather than appetite: shipping companies want onboard AI built on vetted sources rather than the open internet, and most systems still run on role-based accounts that make it hard to track who used AI, and for what. Captains, she added, need confidence that any tool is trustworthy enough to support their decisions before they will rely on it.

Governance gaps are not limited to onboard tools. BIMCO warned in a paper this month that AI adoption deepens shipowners’ dependence on a small number of dominant technology and cloud providers, adding a geopolitical dimension to cyber risk alongside the usual questions of firewalls and patching. Where an AI platform is hosted, and which jurisdiction governs its data, can matter as much as how well it performs.

Put together, shipping is not scaling AI in a total vacuum. There is now a binding international code for autonomous vessels, individual class societies are building their own guardrails, and industry bodies are mapping the new risks that come with dependence on outside AI providers. What is still missing is a single standard covering the broader category of AI decision-support tools now moving from wheelhouse to back office. Until one exists, the industry’s own evidence suggests the limiting factor is not whether seafarers and shore staff want to use AI. It is whether anyone can vouch for what the AI is telling them.

Geneva Dry’s AI, Digitalisation and the Dry Bulk Workforce session, held at the end of April, covered AI and accountability in some detail. 

The session opened with a rapid-fire question to the panel: are shipping companies deploying AI faster than they are defining accountability for it? The room split. Alberto Perez, global head of maritime commercial markets at Lloyd’s Register, and Jonathan Canaan, global ocean freight director at ADM, said yes. Alex Albertini, CEO of Marfin Management, and Ingrid Kylstad, managing director of Klaveness Digital, said no. Scott Bergeron of Oldendorff Carriers offered perhaps the most honest answer of all: “Most of us are probably still trying to figure out how we’re going to deploy AI and not yet worried about the governance of AI.”

Moderator Cynthia Worley of Sedna opened with a warning that landed visibly in the room: the EU AI Act enters full enforcement this year, carrying fines of up to €35m or 7% of global annual turnover for companies unable to demonstrate governance of their AI processes. A show of hands revealed that almost nobody present had heard of it before that week.

Perez outlined Lloyd’s Register’s framework for AI accountability, arguing it must be viewed as part of a wider system rather than in isolation. “The proper definition of accountability needs clarity in three fronts: define the function being deployed, define the decision boundaries of that functionality, and define the performance limits of that deployment,” he said.

The topic will be revisited on September 24 during the AI, Digitalisation, And The Maritime Workforce session at the inaugural Splash Singapore conference taking place at the five-star Fairmont Hotel.