AI for Boards: Practical Advice From What I’ve Seen Work (and Not Work)

In the last few years, I’ve worked with leadership teams who approach AI in very different ways. Some treat it like a mysterious black box, some chase every shiny demo, and a few take a structured, sober approach and actually create value.

If I were to give board members one piece of advice, it’s this: AI is no longer a technology discussion, it’s a leadership and responsibility discussion. And it belongs squarely in the boardroom.

Below are practical recommendations based on what I’ve seen up close.

1. Keep yourself updated but avoid the hype treadmill

You don’t need to understand transformer architectures, but you do need a basic literacy in what AI can and can’t do. That means:

• Follow a few trusted sources, not every headline.

• Understand the difference between productized AI (stable, useful) and experimental AI (impressive, but not always reliable).

• Regularly ask your leadership team: What has changed in AI in the last quarter that affects our business?

This mindset protects you from both naive optimism and unnecessary panic.

2. Treat AI as part of the business strategy, not a side project.

AI is not a “thing the data team does.” It touches:

• customer experience

• operational efficiency

• risk

• talent

• competitive positioning

In other words: AI belongs in the core business plan.

Boards should insist that AI initiatives have the same discipline as any other investment:

• Clear ownership

• Expected business outcomes

• Measurable value

• Sunset criteria if value isn’t materializing

If an AI project cannot explain the business case, it’s not ready.

3. Ask the right questions about threats and opportunities

Every board should regularly challenge management with a simple framework based on opportunities and threats.

Opportunities:

• Where can AI create new revenue?

• Where can AI reduce cost or increase quality?

• Which processes can be redesigned, not just automated?

Threats:

• Could a competitor use AI to undercut our advantage?

• Are we exposed to data, privacy or regulatory risks?

• Are we becoming overly dependent on vendors we don’t understand?

If you’re not discussing both sides, you’re not seeing the full picture.

4. Avoid “AI theatre” and focus on value, not experiments.

I’ve seen companies spend millions on AI without moving a single needle in the business. It happens when:

• experiments run forever

• no one defines success

• teams chase the newest model

• there is no connection to real business processes

Your job as a board member is to demand clarity:

What value have we created with AI this quarter?

How do we know?

What will we stop doing?

Without these questions, you get AI theatre: a lot of activity and very little progress.

5. Make sure the organization is ready, technology alone won’t save anyone.

The biggest blockers I’ve seen are not technical; they are organizational:

• unclear data ownership

• lack of cross-functional collaboration

• slow decision cycles

• fear of changing long-standing processes

• no training for employees

Boards should ensure there’s an AI operating model in place, not just tools. That means skills, policies, ways of working, and clear responsibilities.

6. Set principles for responsible and sustainable AI use

Boards are accountable for risk. That includes:

• transparency around how models are used

• data privacy and governance

• mitigating bias

• monitoring model performance over time

• ensuring human oversight where needed

Responsible AI is not bureaucracy, it’s risk management, and it protects both customers and the company.

7. Build a culture where people are encouraged to use AI

AI succeeds when people use it. That requires:

• psychological safety to experiment

• training for everyone, not just technical teams

• guidance on what tools to use

• sharing success stories internally

I’ve seen organizations unlock massive productivity simply by enabling employees to use AI responsibly in their daily tasks.

8. Don’t try to predict the future, build the capability to adapt.

AI changes monthly. You won’t be able to chart a five-year AI roadmap that survives reality. Instead, focus on:

• a flexible architecture

• fast experimentation cycles

• continuous learning

• a small set of KPIs that track value creation

Boards that require rigid plans end up slowing the organization down.Boards that emphasize adaptability create lasting advantage.

My final thought

AI is not magic, and it’s not optional. It’s a set of tools that, when used responsibly and pragmatically, can fundamentally strengthen a business.

A board’s role is not to be AI experts, but to create the conditions where AI can generate real value, safely and sustainably.

Take a calm, structured, Scandinavian-pragmatic approach:

Less hype.

More clarity.

More value.