I joined the MARE Forum AI and Innovation panel in West Palm Beach last week and prepared this resource for the room. It is republished here for my wider network.
MARE Forum · West Palm Beach · March 2026
AI is advancing faster than most organisations can evaluate it. Over the last 18 months working with maritime companies on AI adoption, I've noticed a few things that keep coming up.
01 — Organisational AI is not a subscription.
Getting a subscription to a consumer chat interface — ChatGPT, Gemini, or any other large language model front-end — is exactly the right place to start as an individual. Learn, experiment, iterate.
At an organisational level, we are doing something different. Think of the intelligence inside a large language model as the kitchen of a food truck. A consumer subscription is the front window — you order, it serves. Organisational AI is the side door. You are connecting that intelligence directly into a flow of tasks, at a completely different price model and a completely different level of integration.
We are not chatting with a large language model. We are plugging it into the way the business works.
The Model Wars — Organisational AI Architecture
Build the system.
The model is interchangeable.
Most conversations about AI stop at the product layer. The work is in the architecture beneath it.
02 — Agentic AI means it takes action — not just answers.
An AI agent is not a fancy search engine. The difference is that it acts on your behalf.
A simple way to understand what an agent does is the DROID framework:
The real value is not one question at a time. It is when you stitch these small processes together into a longer workflow — and decide along the way where a human reviews the workings before it continues. That is where organisations realise significant gains.
A simple agentic workflow in practice. The agent runs without manual input. The output lands structured and ready to act on.
03 — Starting with implementation is not always the right first move.
When you automate processes and speed them up, the quality of the outcome depends entirely on the quality of the foundation underneath. If processes are built on misaligned data or disconnected systems, speed makes that worse, not better. You are not automating efficiency. You are automating chaos.
Most AI initiatives fail.
The reasons are rarely technical.
The vast majority of corporate generative AI pilots are failing to generate meaningful financial returns, despite widespread investment.
MIT — The GenAI Divide, 2025
- Strategic MisalignmentNo clear link between AI initiative and business priorities.
- Poor SequencingRushing to deploy before understanding what should happen first.
- Lack of Executive OwnershipNo leadership accountability for direction or outcomes.
- Weak Workflow IntegrationTools built in isolation from real operational processes.
- No Ongoing DirectionInitial plans abandoned as conditions and technology evolve.
- Blueprint maps prioritiesDocumented direction tied to real operational objectives from day one.
- Sequencing is built inEvery roadmap establishes what happens first — and what does not happen yet.
- Leadership interviews firstExecutive ownership is established before any recommendation is made.
- Workflow & systems mappingEvery recommendation is grounded in real operational processes.
- Navigator preserves alignmentDirection adapts as conditions and technology evolve over time.
In my practice, I always start with a Compass AI Blueprint. It sets direction before anything is built. It maps what AI is in use, what obligations apply — ISM, MLC, GDPR, and whatever is specific to each organisation's subsector — and produces a defensible policy our clients can inspect and our industry can trust.
The governance piece is not a brake on progress. It is what makes progress defensible.
Published Research
AI Governance for the Superyacht Industry — What the frameworks say, how far they reach, and what organisations can do now. First edition, March 2026. Read the Report
If this raised questions for your organisation
The Compass AI Blueprint is the structured starting point for maritime organisations thinking seriously about AI adoption. It maps where you stand, what obligations apply, and what belongs first.
Request the Engagement Guide — No commitment. Review scope, structure, and investment at your own pace.
Kristina Agustin is the founder of Southern Sky AI, a structured AI adoption advisory practice for maritime leaders. southernsky.ai
