Only 14% of companies have clearly defined at leadership level who is actually accountable for AI governance, according to the Logicalis CIO Report 2026. It is a striking number, and also, we think, a slightly misleading one if it is read as a story about neglect. Most of the leadership teams behind that 86% are not ignoring AI governance. Many of them have a programme, a policy, a training rollout, real activity. What they do not have is a named person or team whose job it is to facilitate this.
That distinction matters because activity and accountability are not the same claim, and the gap between them is exactly where the Grant Thornton 2026 AI Impact Survey‘s other figure lands: 78% of business leaders cannot say whether their organisation could pass an independent AI governance audit within ninety days.
Ninety days is not a short window. It is long enough to prepare almost anything, except a named signatory who can point to a decision they changed. Programmes can be assembled quickly. A track record of a team member or team making a call under pressure cannot be assembled at all; it either exists already or it does not.
The two numbers describe the same underlying gap from different angles.
- 14% measures whether accountability has a name attached.
- 78% measures whether that name, if it exists, would survive being checked.
Read together, they say something more specific than “most organisations are behind“: they say most organisations have built the visible parts of governance and left the part that gets tested under scrutiny for later.
Closing that gap is not a matter of writing another policy. It is a matter of naming the person or team, giving them authority, and letting them exercise it before the audit, not during it…