Should we appoint a head of AI?

Faced with AI, the first reflex is often to create a post: a head of AI, sometimes a ‘Chief AI Officer’, in charge of steering the subject. It is reassuring: the subject finally has a name and an address.

The risk is isolating that person. Without authority over the business or over budgets, they facilitate, take stock and advise, but do not decide. The real trade-offs, what gets automated, what is kept, what is stopped, are still made elsewhere, or not made at all.

• The subject drifts towards technology.A head of AI without a strong link to the leadership often ends up reporting to IT, and AI becomes a matter of tools.
• The business waits for solutions to be brought to it.Business teams feel consulted rather than responsible, and commit less.
• Difficult decisions are escalated without being settled.Choosing one use over another commits the company: that falls to the people who run it.

1. 01Name the member of the leadership team who will own AI, and make it known.
2. 02If you create a coordination role, have it report to that person rather than to a department.
3. 03Put decisions about AI on the agenda of the bodies that already set the company’s priorities.

We help leadership teams organise who decides what, before the first projects are launched.

• Who should govern AI in the company?→
• Who should lead the project: IT or the business?→
• ‘AI is an IT project.’→
• We already have ChatGPT or Copilot licences. Is that an AI strategy?→