AI adoption in shipping raises a structural question about expertise formation
As automation handles more routine judgment calls at sea, the industry is asking how the next generation builds the experience base that AI was trained on.
THE NEWS
According to Splash247, the maritime industry's conversation around artificial intelligence is shifting away from technology capability and toward workforce consequences. The pivot was on display at the AI, Digitalisation and the Maritime Workforce panel held at Splash Singapore, where executives drawn from shipping, technology, classification, and mining engaged with the tension between AI-assisted operations and the development of human expertise. The central concern, as framed by the panel, is what happens to the people expected to work alongside these systems.
The discussion surfaced a structural challenge that goes beyond job displacement: if AI systems absorb the routine decision-making that has historically served as the training ground for junior and mid-career professionals, the pipeline of experienced practitioners may narrow over time. Panelists from across sectors — not just shipping — appear to have converged on this as a shared concern rather than a problem unique to any single industry.
The article does not detail specific policy proposals or technical frameworks that emerged from the panel, but the framing itself — expertise formation rather than headcount — marks a meaningful evolution in how the sector is publicly discussing automation.
WHY IT MATTERS
For the Brazilian offshore sector, this debate arrives at a moment of considerable operational scale. Brazil's deepwater and ultra-deepwater activity involves a dense ecosystem of vessels, subsea systems, and support infrastructure, all of which depend on a continuous supply of technically proficient professionals — marine officers, subsea engineers, ROV pilots, dynamic positioning operators, and the supervisory layers above them. The question raised at Splash Singapore is directly applicable: how do those professionals develop the judgment that AI tools are now being asked to replicate or assist?
The concern is structural rather than immediate. AI systems in maritime contexts — whether for vessel routing, predictive maintenance, or operational decision support — are typically trained on historical data generated by human practitioners making real decisions under real conditions. If those practitioners are progressively removed from the decision loop earlier in their careers, the experiential data that future AI systems would need to improve becomes thinner. The industry could face a compounding dynamic where automation reduces the formation of expertise, and reduced expertise constrains the quality of future automation.
Brazilian operators and their contracted vessel managers face a version of this challenge that is shaped by local conditions. The Brazilian maritime workforce operates under a regulatory framework that mandates significant national crew content on vessels working in Brazilian waters. This means that the question of expertise formation is not purely a global abstraction — it has direct implications for how Brazilian maritime academies, ANTAQ-regulated operators, and offshore service companies structure their training pipelines. If the tasks that once built competence are increasingly handled by software, the curriculum and at-sea experience requirements that underpin Brazilian maritime certification may need to be revisited.
For Petrobras and the independent operators active in the pre-sal and post-sal plays, the workforce dimension of AI adoption sits alongside the more visible capital allocation and technology integration questions. Vessel operators providing support to FPSO operations, for instance, rely on DP officers and marine superintendents whose expertise is accumulated incrementally over years of hands-on exposure. A DP incident in ultra-deepwater is not a recoverable situation in the same way a software error onshore might be. The margin for experiential deficit is narrow, and that asymmetry gives the expertise-formation argument particular weight in the offshore context.
The classification society presence on the Splash Singapore panel is also worth noting analytically. Classification bodies set the technical standards against which vessels, systems, and increasingly software tools are assessed. Their engagement with the workforce dimension of AI suggests that the regulatory and standards community is beginning to think about how human competency requirements should be defined in an environment where AI handles more of the operational load. For Brazilian operators working with international class societies, any evolution in competency standards will have direct compliance and crewing implications.
CONTEXT
The expertise-formation concern is not unique to maritime. Aviation, nuclear operations, and anaesthesiology have all navigated versions of the same tension as automation absorbed tasks that previously built practitioner judgment. In each case, the resolution involved deliberate design of training environments that preserved exposure to consequential decisions — simulators, graduated responsibility structures, and mandatory minimums for unassisted operation. Maritime regulators and operators are likely to draw on those precedents as the AI integration debate matures.
Brazil's offshore sector has navigated significant workforce transitions before, including the rapid scaling of pre-sal operations and the localization requirements that shaped how international contractors staff their Brazilian projects. The current AI moment is a different kind of transition — slower, less visible, and harder to regulate through conventional headcount rules — but the institutional capacity to manage workforce change at scale is present in the ecosystem.
Source: SPLASH247