AI strategy and AI governance

What AI-mature organisations do differently

Almost everyone is experimenting with AI by now. The gap opens up at scale, and that turns out to be mostly an organisational question.

· · 3 min read

For its AI Maturity Index 2025, IMD studied the 300 largest companies in the Forbes Global 2000. The researchers sum up their findings as a matter of alignment: technology that fits the governance, partners that fit the goals, and people with the skills the ambition requires. That sounds obvious. Most companies still struggle with it.

The hundred most AI-mature companies also grew faster in revenue and market value. That is a correlation. Whether AI is the cause, or well-run companies simply happen to be better at AI as well, the research cannot say. I read it as a signal for leadership teams rather than as proof.

What surprised me most: financial services has 32 companies in the top 100, more than the technology sector itself.

What the leaders have in common

The clearest difference is where AI sits in the organisation. Among the leaders it sits at the leadership table, with a clear owner or an AI council steering adoption across functions. An innovation lab on the sidelines is rare in the top 100.

They also invest heavily in people. According to the research, Unilever trained more than 23,000 employees in AI skills and Visa reskilled 10,000. They publish their AI principles and set up oversight, because they see it as a condition for trust as well as a compliance matter. They build far from everything themselves and work closely with technology partners. And they increasingly use AI for growth, through new products and business models, where it used to be mainly about automation.

Data first, but targeted

Watkins and Weissbeck name seven foundations for successful AI: data, processes, technology, governance, talent, training and leadership. Their first piece of advice is the least glamorous: do the dirty work on your data first. They attach a condition that often gets skipped. Choose the highest-value use cases first and only then start cleaning. Start the other way round and the budget disappears into a data clean-up with no direction.

Manufacturing shows the same pattern. The industrial leaders combine domain knowledge with their own data, use AI in maintenance and quality, and invest systematically in their people. There, AI is simply part of how the factory runs.

Three questions for the leadership team

Gopi Kallayil, AI strategist at Google, suggests three questions, in this order: what is our growth plan for this year, what are the priorities and the biggest challenges, and how can AI enable that strategy? That AI only comes up in the third question is exactly the point.

What this means in practice

In my experience, an AI programme rarely stalls because of the technology. More often nobody is clearly the owner, the data is not in order, or no one knows what success looks like. If I could give one piece of advice: get an owner at the leadership table first. Then pick a few use cases with a business case and make sure the data can carry them.

Sources

  1. Yokoi, T., Wade, M. & Shan, J. (2025). IMD AI Maturity Index 2025: AI strategies that are working. IMD, TONOMUS Global Center for Digital and AI Transformation.
  2. Yokoi, T. & Wade, M. (2025). AI maturity in manufacturing: lessons from the most successful firms. I by IMD, 9 August 2025.
  3. Watkins, M. D. & Weissbeck, R. (2023). The 7 foundations of successful AI implementation. I by IMD, 28 September 2023.
  4. Kallayil, G. (2025). How can AI enable your business strategy? I by IMD, 1 July 2025.

Read next

Let's talk

Would you like to exchange ideas on data, analytics, AI or FP&A? I would be glad to connect.