A recent study by Thetius in partnership with Marcura, has revealed maritime companies are stuck in the early stages of AI adoption, unable to scale beyond small experiments as widespread optimism collides with implementation reality.
The study “Beyond the Hype: What the maritime industry really thinks about AI” combined over 130 survey responses and in-depth interviews with maritime professionals, revealing a sector that is both eager and cautious: 82% are optimistic about AI and 81% are running pilot projects. However, 37% have personally witnessed AI failures and only 11% have formal policies to guide scaling.
Perceived risks and opportunities
- 97% believe that AI is useful or extremely useful for reducing manual workflow inefficiencies
- 85% say that AI is useful or extremely useful for identifying risky voyage decisions and red flags in voyage profitability
- 69% are concerned about poor business outcomes if AI solutions miss critical red flags in contracts or voyage planning
- 66% worry that overreliance on the technology could lead to a reduction in human skills and oversight
- 61% feel that cybersecurity and data breach vulnerabilities are the biggest risks for them in implementing AI in maritime operations
- 37% have witnessed AI projects failing or causing harm
- 23% feel that vendors are generally untrustworthy, offering too much hype and insufficient results.
According to the study despite their general enthusiasm for AI, maritime professionals overwhelmingly reject full automation. As explained, 70% believe AI should recommend actions but humans should always make the final decision, while 66% are concerned about overreliance on the technology eroding human skills and judgement.
Additionally, the study identified inadequate training as the biggest barrier to scaling, cited by 38% of respondents. The governance gap is equally stark: while 81% run pilots, only 17% have transparent processes for how AI makes decisions within their organisations. Nearly a quarter express concerns about vendor claims outpacing real-world results. Concerns such as data privacy and cybersecurity are also significant, with 61% citing them as major risks. Strong data governance and ownership frameworks are needed before organisations feel ready to scale. High- quality data is also critical to success.
The rapid shift in AI adoption
Companies are using AI in areas such as navigation automation and cargo operations. But according to Theofano Somaripa, CIO at Newport SA, many smaller and medium- sized ones are not yet ready to fully adopt or scale AI. According to the research, financial constraints, lack of digital transformation strategies, staff readiness, and concerns over data privacy and transparency are blockers to scaling AI.
One of the reasons for maritime’s rapid adoption of AI when historically it has been fairly slow to embrace new technologies, is the increasing pressure for companies to show they are committed to advancing technologies and keeping pace with industry leaders.
Another reason for the accelerated maturity curve is due to vendor relationships. Companies that have used solutions from one vendor may be more inclined to adopt AI enhancements built on top of those foundations. This is because they have not only built trust with that vendor but also because their data is already digitised and adding AI is the next natural step.
Furthermore, vertical AI is building trust and maturity faster. Due to it being built for the industry’s specific needs, it delivers faster time-to-value, enabling companies to progress from pilots to deployment more quickly and with greater confidence.
“The best AI functions like a co- pilot, not a replacement, providing insights but always leaving the final decision up to the professional who understands the full context. As seen in some legal cases, relying on AI without human oversight can cause errors and even cross into professional misconduct,” said Janani Yagnamurthy, VP Analytics, Marcura.
Key Recommendations
- Invest in tools specifically for maritime
- Foster agency and discernment
- Keep the human in the loop to harness trust
- Engage with emotions, not just system
- Implement governance frameworks
- Demand transparency and real-world impact from vendors
- Encourage experimentation