This insight outlines a practical framework for implementing AI in maritime operations in a way that strengthens human decision-making and maintains safety, accountability, and trust.
A framework for AI-human collaboration is required to ensure that the use of AI is safe, efficient, and ethically sound. Before AI can be used effectively in voyage optimisation or operational decision-making, the underlying data must be cleaned, validated, and interpreted.
As Manuel Fernandez de Mosteyrin, Manager Chartering, Products & Chemicals at Stena Bulk, explained, “Most problems are data problems, not AI problems.” The real effort lies in cleaning and structuring data, aligning formats and taxonomies, fixing integrations between systems, and establishing clear governance so decisions are not made in silos.
Data must also be continuously monitored for drift or degradation, with clear processes for humans to review suspect inputs or outputs. Without strong data foundations, even the most sophisticated optimisation logic will generate unrealistic or unsafe results.
Balanced AI is not engineered in isolation; it is co-designed with the people who rely on it. Users must be integrated throughout the lifecycle, from design and development through testing, deployment, and operation. This ensures AI reflects maritime reality rather than generic assumptions.
Explainability, transparency, and traceability are essential. Users must understand why a route or speed was selected, what trade-offs were made, and under what constraints. Without this, AI becomes a black box and loses operational credibility.
KPIs must also be redesigned for AI-enabled workflows. Existing KPIs measure manual work rather than how effectively people use, challenge, or supervise AI systems. Strong user interface design is equally critical, acting as the bridge that makes AI usable, understandable, and trusted.
Finally, organisations should establish certified, contained AI environments, supported by governance structures that define accountability, safety thresholds, and exceptional situations. Continuous learning and human oversight ensure AI remains accurate, ethical, and resilient as operations evolve.
To explore how a balanced human-AI approach can strengthen voyage optimisation in safety-critical conditions, read the full Thought Leadership report ‘Co-Pilots of the Sea: Exploring the Human Intelligence Behind Maritime AI’ produced in partnership with StormGeo.
