Episode 4, Part 3: Considerations for AI Adoption — This episode explores the critical considerations for adopting AI in maritime businesses: security, responsible use, and environmental sustainability.
Security and Confidentiality
As we bring AI into our businesses, we need to do it in a way that reflects our values—using it securely, responsibly, and sustainably. In our industry, confidentiality is foundational.
We handle ultra-high-net-worth (UHNW) clients, sensitive crew data, and commercially valuable business information. How we engage with AI—where data goes and how it's stored—are critical decisions.
"Fear of data risk often holds businesses back, but we can manage those risks and still take advantage of AI's value."
AI is already used in some of the world's most secure environments, including military and government.
The AI Security Spectrum
A practical spectrum of ways to engage with AI while keeping privacy and data security in mind—from Level 1 (free tools like ChatGPT) to Level 6 (fully isolated, top-secret systems).
Buying AI Tools
- —Level 1 – Free Public Tools: Tools like ChatGPT should never be used for confidential data. They often train on your inputs, which can end up in public responses
- —Level 2 – Paid Tools: ChatGPT Pro or Business offer better data handling, but they're still public platforms. Best for low-risk use
- —Level 2.5 – Professional SaaS Platforms: Purpose-built with clear data policies and certifications, making them safer for business use. Always check how your data is managed and whether tools train on your data
- —Level 3 – Enterprise AI Tools: Tools inside your infrastructure offer stronger data control, but may not currently match the performance of leading open-source models (e.g., Microsoft Copilot inside Word)
Building AI Systems
- —Level 4 – Private Cloud AI: Uses platforms like AWS, giving you strong security without owning the infrastructure
- —Level 5 – Fully On-Premise AI: Runs on your own servers, keeping data entirely in-house, but with high costs and complexity
- —Level 6 – Fully Air-Gapped AI: Used for highly classified environments—wonderful to have, but overkill for most business needs
The Sweet Spot
Most superyacht businesses will find the sweet spot between Level 2.5 (SaaS tools) and Level 4 (private cloud AI systems). These levels give you powerful AI while keeping data protected without heavy infrastructure.
For highly sensitive data, you can use a sanitise, interact, and re-identify approach: strip out sensitive details locally, use the AI model, then reinsert the data back inside your secure environment.
Responsible AI
Because AI models are trained on the entirety of human history, they are by nature shaped by our past. And our past hasn't always been fair.
History carries deep-rooted biases across gender, race, class, and geography. AI systems will inevitably carry forward some of those same biases unless we deliberately intervene.
Key Principles
- —AI can't genuinely reason and lacks human empathy
- —Regularly audit outputs and monitor for bias
- —Important decisions—especially those affecting people's lives—must have human oversight
- —The principle of human-in-the-loop is essential for transparency and accountability
"For most businesses, you'll be using AI models, not building them. As users, you are always responsible for the outputs and what happens as a result of them."
Environmental Impact
Given our industry's commitment to sustainability, we must consider AI's environmental impact.
The Challenge
AI's energy use is rising fast. By 2030, data centres could consume 4.5% of global electricity—equivalent to multiple nuclear plants.
Industry Response
Big tech is acting. Microsoft, Google, and AWS have set aggressive targets: carbon negative, water positive, zero waste by 2030. Google is also taking steps to bring nuclear energy into the AI supply chain.
What You Can Do
At a business level, you can reduce impact by:
- —Using smaller models for day-to-day tasks
- —Right-sizing computing power
- —Batching processes where possible
- —Keeping prompts focused and workflows efficient
AI for Sustainability
Importantly, AI can drive sustainability too. It can help monitor, predict, and optimise operations—cutting resource use and emissions.
The Rossinavi SeaWolf X Catamaran, launched in May 2024, is a great example where AI reduces fuel use and emissions across the vessel's lifecycle.
"I see the current movement in AI being the driver that pushes our society into clean energy adoption, fuelled by the movement of big tech."