Who’s At The Helm? Bridge Discipline In Maritime AI

By Michelle Ramsden, Ilsa Luther and Andy Lee, Jones Walker LLP

Artificial intelligence (AI) is steadily transforming the maritime industry’s architecture, from predictive maintenance, safety and ship security to navigation and port logistics, as well as the development of autonomous vessels. Even the inland barge and transportation industry, a sector traditionally characterized by manual processes, legacy systems and narrow operating margins, is experiencing this shift. As operators face increasing pressure to improve safety, efficiency and environmental performance along major waterways, such as the Mississippi River system, AI-driven technologies are emerging as practical tools to modernize fleet operations and logistics.

Marine insurers have already begun to recognize the industry’s readiness for innovation and are collaborating to drive efficiencies in pricing, risk assessment and claims handling. Much like a vessel itself, AI requires a level of “bridge discipline,” the same operational rigor and decision-making expectations that govern vessel command, whether offshore or along inland waterways, to ensure it operates safely, remains accountable and stays on course toward reliable and responsible outcomes in complex, real-world environments.

The National Association of Manufacturers reports that AI already provides a critical check on shipping agents’ misdeclaration of hazardous cargo, which has been linked to a concerning increase in deadly fires. Examples include the 2019 fire on the Yantian Express, in which “coconut charcoal,” labeled as “coconut pellets,” ignited, and the 2025 fire on the Wan Hai 503, which killed four crew members off the coast of India. For inland terminals, the potential damage from mis-declared cargo may be even greater, given their proximity to population-dense areas. At the same time, machines are not necessarily more reliable or more straightforward than human operators.

Joanne Waters of the UK’s DAC Beachcroft explains that, for her clients, accountability in a partially and increasingly autonomous environment is more difficult to assign, particularly under pressure. It becomes harder to determine when, and whether, a human should intervene to course-correct systems designed to reduce human error and that lack instinct developed through lived experience. Overlapping jurisdictions can further create regulatory conflict and confusion in the absence of contracts that clearly allocate liability. AI-designed maritime routes can also give rise to intellectual property disputes affecting operators and shipowners.

Innovation Cycles

These trends are not new to technology attorneys. Innovation often outpaces oversight in a predictable cycle: technological advances automate and accelerate labor-intensive processes, expectations rise, and gaps in governance trigger concerns about market share and profitability. AI governance, however, is not just a safety check. It is bridge discipline. It is the professional rigor required to deliver on the promise of advanced technology investment.

The challenges of AI regulation and overlapping jurisdictions are layered. The maritime industry, like the technology sector, has long operated across competing global economies and governance frameworks. While there has been some regulatory retrenchment, industries are not left entirely without guidance. Durable global standards and governance frameworks continue to provide a reliable baseline for organizations deploying AI.

Bridge Discipline Frameworks

Translating bridge discipline into practice requires structured governance, and existing frameworks offer a useful starting point. The National Institute of Standards and Technology AI Risk Management Framework (RMF) is among the most widely adopted and well-resourced baselines for maritime AI. At its core, the RMF organizes governance into four interrelated functions: govern, map, measure and manage.

“Govern” requires organizations to clearly define their intent and boundaries. Key questions include how AI will be used and, more importantly, what uses are considered permissible. Applications may include monitoring maintenance needs, identifying hazardous cargo, predicting security threats, planning efficient shipping routes or automating container handling and scheduling. Organizations must also determine which uses fall outside their risk tolerance. Ethical frameworks and a clear understanding of risk tolerance are essential to responsible innovation.

“Map” requires clarity around accountability structures. Maritime operations rely on a network of complex, interrelated contracts. Organizations implementing AI should ensure that agreements with vendors and downstream counterparties address risks such as bias, data exposure and model error and clearly allocate liability. Internal governance structures should also be established to review and mitigate AI-related risks. Critically, organizations must identify who has authority to disable or intervene in the event of malfunctioning systems.

“Measure” focuses on an organization’s ability to explain and validate AI-driven outcomes. Insurers, regulators and courts will expect organizations to understand how and why an AI system made particular decisions. Ongoing monitoring is essential and may itself be automated to detect issues such as model drift and data degradation.

“Manage” addresses how an organization responds when something goes wrong. In practice, this is a matter of timing rather than possibility. UK-based maritime cybersecurity firm Cydrome reports that up to 60% of newly disclosed software vulnerabilities across maritime environments are weaponized within 48 hours, as threat actors increasingly use AI to accelerate attacks. Beyond external threats, organizations must define how internal anomalies are escalated and when they must be reported under regulatory or contractual obligations. Personnel should be trained on appropriate interventions, and organizations must maintain documentation protocols to mitigate liability and protect against reputational harm.

IMO Instruments

The International Maritime Organization’s Maritime Safety Committee has adopted industry-led instruments supporting autonomous vessel operations within the current regulatory framework. MSC.1/Circ.1638 outlines the results of a regulatory scoping exercise for Maritime Autonomous Surface Ships, identifying areas requiring clarification or revision. MSC.428(98) addresses cybersecurity by incorporating cyberrisk management into ship and company safety management systems.

AI presents significant opportunity across the maritime sector, including inland waterways. However, responsible parties must reclaim the helm, not only to adopt AI but also to remain accountable for it when it fails or deviates from course. Conducting a “bridge audit” of organizational AI use and governance structures can help establish baseline standards, prepare organizations for evolving regulation and protect against reputational harm that can undermine the value of innovation.

(Michelle Ramsden is special counsel in the Corporate Practice Group at Jones Walker and a member of the commercial transactions and privacy, data strategy and artificial intelligence teams. She brings more than a decade of experience leading comprehensive privacy, cybersecurity and artificial intelligence (AI) programs for complex organizations.

Ilsa Luther is an associate in the Maritime Practice Group, a member of the maritime transactional team and of the firm’s Energy and Natural Resources Industry Team.

Andy Lee is a partner in the Litigation Practice Group and a member of the corporate compliance group. He founded and serves as co-leader of the firm’s privacy, data strategy and artificial intelligence team and holds the CIPP/US designation from the International Association of Privacy Professionals. He also chairs the firm’s Innovation Committee and co-leads the Technology Industry Team.)