AI for passenger vessel operators.

The DSM Certificate runs five years, but the system runs on your own paper between audits, the mid-term visit and the annual self-assessment, and 38.3 million domestic sea passengers travelled in 2024, most of them on river ferries and island runs. What AI is already doing across operations like yours, and the safest way to start.

Prefer to talk it through first? Book an introductory conversation.

of tourism operators reporting AI already in use across parts of the business.

OTA commission on every booking before processing and cancellation costs.

Those same uses form a pattern without an owner: guest records moving through tools that were never procured, safety-adjacent documents drafted by systems that have never stood on your deck, and a brand that carries the story if one of them gets it wrong. You already run the discipline this needs. Your safety management system names who checks what before the vessel leaves the berth, and AI governance applies the same discipline to a new hazard class, so the productivity your teams found stays and the practice becomes reviewable.

Figures as published on southernsky.ai, reviewed September 2026.

At the ferry tier the deployment is public: Stena Line runs an AI chatbot for UK customers, Direct Ferries has put ferry search inside ChatGPT across more than 4,000 routes, and UK-based vendors market AI navigation and risk systems to ferry operators. Below that tier AI arrives the way it arrives everywhere, through the booking stack and the office rather than through a decision: platform agents answering guests about refunds, inclusions and weather policy from live availability, crews drafting review replies and campaign copy in public tools, supervisors polishing incident reports before they enter the record.

The UK asks for none of this by statute. There is no AI act; government policy runs five principles through the regulators you already answer to, with UK GDPR over passenger and marketing data, the ICO, the UK's data protection regulator, holding the operative AI guidance, and its statutory code of practice on AI and automated decision-making in progress. What endures is evidentiary: the risk assessment the MAIB reads, the self-assessment the MCA reads, and the refund promise your booking agent makes read under consumer law with turnover-percentage penalties behind it. The workflows stay; a named person checks what reaches a guest, the crew or the record.

Crews draft campaign copy and guest-review replies in ChatGPT and Canva, turning an evening job into a short one, and a named person reads each post before it goes out in the operator's name.

Guest answers on the booking stack

Platform agents such as FareHarbor Agent and Rezdy assistants answer guests about refunds, inclusions and weather policy from live availability, so the register names who owns and checks what those agents say.

Incident reports and toolbox notes

Supervisors polish incident reports and toolbox-talk notes with a chatbot, and FareHarbor's own survey lists SOPs, agreements and training materials among operator uses, so a named person verifies the wording before it reaches crew or the regulator.

Owners draft grant applications and funding submissions with AI, and the operator confirms the figures and the claims before lodging.

Rezdy's pricing engine proposes fares against demand, and a person signs the rate card before it goes live.

These are the workflows the prompt library and training stand up, under the standard the AI use policy, approved-tools register and verification checkpoints set.

• 01Marketing and reviewsCrews draft campaign copy and guest-review replies in ChatGPT and Canva, turning an evening job into a short one, and a named person reads each post before it goes out in the operator's name.
• 02Guest answers on the booking stackPlatform agents such as FareHarbor Agent and Rezdy assistants answer guests about refunds, inclusions and weather policy from live availability, so the register names who owns and checks what those agents say.
• 03Incident reports and toolbox notesSupervisors polish incident reports and toolbox-talk notes with a chatbot, and FareHarbor's own survey lists SOPs, agreements and training materials among operator uses, so a named person verifies the wording before it reaches crew or the regulator.
• 04Grant and funding writingOwners draft grant applications and funding submissions with AI, and the operator confirms the figures and the claims before lodging.
• 05Dynamic pricingRezdy's pricing engine proposes fares against demand, and a person signs the rate card before it goes live.
• 06These are the workflows the prompt library and training stand up, under the standard the AI use policy, approved-tools register and verification checkpoints set.

Figures as published on southernsky.ai, reviewed September 2026.

Founder, Southern Sky AI

20+ years in international superyacht and maritime operations. Admitted Lawyer (Supreme Court of NSW). ATSE Elevate Scholar 2026, completing a Master of Artificial Intelligence.

Figures as published on southernsky.ai, reviewed September 2026.