Experiences and Reflections from a “Non-Expert” in AI
After 20 years of board experience, I’ve seen how important it is for boards to have a holistic view. Strategic discussions should focus less on yesterday’s reports and more on tomorrow’s opportunities and risks. I’ve now explored how AI can be a helpful tool for doing just that.
Recently, I attended the INSEAD Advanced Board Programme, where one of the topics was the use of AI in the boardroom. I don’t believe AI will replace human decision-making anytime soon, but it can undoubtedly enhance the quality of board discussions. Some boards are likely already using AI actively. However, my impression is that AI is mainly applied in the operational work within companies, not in the board’s own processes. At INSEAD, we experimented with using a “chatbot” in simulated board meetings, asking it for input to discussions. We also explored how AI can support the preparation work each board member does before meetings. Both exercises proved valuable to the quality of our discussions.
Since then, I’ve reflected on how this might apply to my own board work. What has particularly inspired me is how insightful and efficient AI can be for testing strategic hypotheses and scenarios. I experimented with formulating possible scenarios and, with AI’s help, received surprisingly solid discussion material about their content and implications. By using AI as a sparring partner and dialogue tool, rather than just for simple queries, I gained meaningful insights with very limited time investment. Importantly, this could be done without revealing any business-sensitive information — something that matters to most boards.
Most boards do not have specific AI competence among their members. Nor is it necessarily right or feasible to recruit an AI expert. What matters most, in my view, is that board members are open to exploring AI as part of their governance work and start testing it. It has become very easy to “play” and spar with, and what you get in return from using AI is already useful — and improving rapidly. It might even make sense to include it as a standing item on the board agenda, for example: “How can we use AI to improve our board work in this meeting?”
Below, I share an example of my own scenario testing and how I believe AI can help boards work with future opportunities and risks. I used the shipping industry as my starting point — an industry I know well, characterized by complex and cyclical markets. I proposed three scenarios and received surprisingly strong analyses of their complexity, consequences, and the strategic considerations a board should address. I also asked AI to suggest one additional scenario, and it produced a relevant and thought-provoking alternative. AI did not cover every aspect or deliver flawless answers, but I can easily see how this could serve as a strong foundation for boardroom discussions.
I don’t believe we will let AI make decisions for us just yet, but the use of AI will certainly change how we work toward decisions. It can help us discuss smarter, think further ahead, and act faster. The value of scenario testing, for example, lies in exploring what could happen, assessing the consequences of different events, and emphasizing the need for flexibility — both in mindset and business model. The likelihood of a “black swan” event has increased significantly in today’s complex world. It didn’t take much time or experimentation to confirm that using AI in the boardroom provides a stronger foundation for dialogue and strategic decision-making — and I recommend everyone who hasn’t yet tried it to do so.
Scenarios – an Example of My Dialogue with AI on Scenario Analysis in the Shipping Industry
Below are three scenarios I asked an AI chatbot to elaborate on for strategic board discussions in a shipping company, along with a request for a fourth alternative. Each scenario outlines assumptions, possible consequences for shipping, and strategic reflection points for the board to consider.
⚓️ Scenario 1:
Steady State – “Business as Usual”
Description
The global economy develops in line with trends from the past decade. Global trade grows moderately (2–3% annually), the energy transition progresses gradually, and geopolitics remain manageable. Technological development and regulations (ESG, digitalization) proceed at an expected pace.
Assumptions
- Freight volumes increase in line with GDP.
- China and other Asian shipyards expand moderately, avoiding oversupply.
- Fuel and energy costs remain stable.
- Emission and reporting regulations (EU ETS, IMO 2030) evolve gradually and predictably.
Consequences for Shipping
- Relatively stable rates, but no super-profits.
- Investment decisions can be made with a long-term perspective.
- Digitalization and AI can improve efficiency but are not existentially necessary.
Strategic Reflections
- How can AI and data analytics be used for continuous operational and fleet optimization?
- How should investments be balanced between green technology and traditional fleet renewal?
- What is the risk of becoming too “comfortable” in a stable environment?
⚙️ Scenario 2:
Oversupply on the Supply Side – “China Floods the Market”
Description
China invests heavily in shipbuilding as an industrial growth driver, supported by subsidies and automation. This leads to a sharp increase in global shipyard capacity, and newbuilds are aggressively priced. Meanwhile, freight demand grows only moderately.
Assumptions
- 15–25% growth in global fleet capacity over 5–7 years.
- Lower newbuild prices reduce capital barriers.
- Many players order new “green” ships to meet future regulations — but this further amplifies overcapacity.
- Freight rates come under strong pressure.
Consequences for Shipping
- Margin pressure and lower returns.
- Increased risk of restructurings and consolidations.
- Growing gap between efficient (digitalized, AI-driven) operators and smaller traditional ones.
- Stronger need for differentiation — e.g., sustainability, data-driven operations, partnerships, or integrated logistics.
Strategic Reflections
- How can AI be used to forecast rate developments and capacity balance?
- Should the company adjust its investment strategy — e.g., focus on niche markets or long-term contracts?
- Is it time for strategic alliances or mergers to strengthen the position?
- How to assess technological risk in newbuilds (e.g., green fuel technology not yet mature)?
⚔️ Scenario 3:
Hybrid Warfare and Energy Instability – “A Fragmented World”
Description
Geopolitical unrest affects large parts of the world. Cyberattacks, drone threats, and unstable energy supply cause fragmented value chains. Some regions experience prolonged energy shortages. Transport and logistics systems become vulnerable, increasing the strategic importance — but also operational risk — of maritime transport.
Assumptions
- Energy access is uneven, and fuel prices fluctuate sharply.
- Digital infrastructure (AI, sensors, communication) becomes a target for attacks or instability.
- Political blocs shape trade and shipping routes (regionalization, sanctions, blockades).
- Rising demand for “resilient logistics” — secure, flexible transport chains.
Consequences for Shipping
- Highly unpredictable rates and trade patterns.
- Increased insurance and security costs.
- Greater value in ships and companies offering autonomy, flexibility, and data security.
- High demand for operators able to reroute fleets quickly and operate safely under uncertainty.
Strategic Reflections
- How can AI support risk modeling, route optimization, and real-time readiness?
- Should the company invest in its own energy security (e.g., hybrid, hydrogen, or battery systems)?
- How to address cybersecurity and protection of digital systems?
- Can the company position itself as a “partner for resilient supply” in turbulent times?
💡 Summary – AI as a Strategic Tool in Scenario Thinking
AI can contribute across all scenarios by enabling:
- Scenario testing: Simulating market shifts, cost levels, and freight rates.
- Real-time risk analysis: Monitoring geopolitical, economic, and technological signals.
- Decision support: Generating “what-if” analyses faster than traditional methods.
- Improved board discussions: Visualizing complex interconnections in a clear way, helping the board focus on strategy — not just details.
Scenario 4:
Green Acceleration – “Capital and Customers Go Green Faster than Regulators”
Description
Demand for sustainable solutions accelerates, driven by investors, insurers, and major charterers setting their own climate requirements. Capital flows away from companies unable to document emissions reductions, and green financing becomes the norm. Technological development in alternative fuels (ammonia, hydrogen, methanol) and autonomous systems advances faster than expected.
Assumptions
- New climate and ESG requirements make traditional vessels obsolete quickly.
- Green technology becomes commercially viable sooner than anticipated.
- Customers — especially in energy, bulk, and container sectors — demand low-emission shipping and long-term green contracts.
- Capital markets reward verifiable sustainability and transparency.
Consequences for Shipping
- Access to capital and insurance terms increasingly tied to emission profiles.
- High technological risk — choosing the wrong fuel technology could cause losses.
- New partnerships emerge between shipping companies, tech firms, and energy producers.
- AI becomes key to simulating emissions, optimizing fuel choices, and documenting sustainability to clients and investors.
Strategic Reflections
- How to balance green investments against the risk of technological missteps?
- Can the company take a “first mover” position in green shipping — or is it safer to wait?
- How can AI be used to simulate carbon costs and future ESG requirements?
This scenario complements the previous three:
- Steady State (continuity)
- Oversupply (market disruption)
- Hybrid War (geopolitical unrest)
- Green Acceleration (regulatory and technological disruption)
This was the result when I provided AI with three scenarios and asked for one more. The output looks solid and reasonably sound — though not perfect. An industry expert, for example, might note that in Scenario 2, automating ship welding is still unsolved, and the shipbuilding industry faces a shortage of welders. In Scenario 4, one could argue that it overlooks the critical issue that everything becomes more expensive — and ultimately, consumers must pay the price. The question then is whether society and consumers are truly willing to do so. It also doesn’t address energy security; nations are deeply concerned about securing energy access, and the green transition isn’t progressing fast enough to guarantee it.
These are the kinds of discussions that can continue in the boardroom — and still be explored with the chatbot. Even so, it provides a strong foundation for discussing different scenarios without spending excessive time preparing them.
It’s also worth noting that AI relies on historical data and context. This can be addressed by developing an internal AI system for use in board and management work — one that doesn’t share data externally and can be tailored to the company’s needs. Of course, this requires resources and expertise, but I’d argue that boards benefit from experimenting with existing AI tools before considering building their own.
Ultimately, what matters most is that board members are open to using AI in their work and start testing it. It has become very easy to use, and what you gain from it is already valuable — and constantly improving. From there, boards can decide how to take it further.