Updated 17 October 2026. This article has been revised to separate the policy timeline from the product argument. See the current AI Act timeline for the category-specific dates and primary reference.
A calendar can create urgency. It cannot assign responsibility. When the date changes, a team with no clear owner still has no clear owner, only a newer calendar invitation.
Give the work an accountable owner
Imagine a system assessment that requires product context, legal interpretation and engineering evidence. Giving one person responsibility does not mean asking that person to do all three jobs. It means somebody is responsible for bringing the answers together and exposing unresolved disagreements.
A useful ownership record identifies the decision-maker, the people providing evidence and the person who can block release when an agreed condition is not met. Avoid a committee-shaped blank space. If every department is responsible in exactly the same way, nobody knows who must act next.
The owner also needs authority proportional to the responsibility. A person cannot credibly own a risk if they cannot obtain the relevant records, request a test or escalate an unacceptable result. Accountability without access is mostly a particularly unpleasant job title.
| Decision | Evidence to request |
|---|---|
| Assign ownership | Named person and decision scope |
| Request evidence | Access to records and test owners |
| Escalate uncertainty | A route to someone authorised to decide |
This is not dysfunction in the classic sense. It is a rational response to a governance structure with no center of gravity. When nobody is clearly accountable, everyone is cautiously accountable, which is functionally the same as nobody being accountable at all.
The rest will use this extension the same way most people use a snooze button. It feels like rest. It is not.
What to do with this
Try the arrangement on one unresolved issue. Who requests the missing evidence? Who decides whether the remaining uncertainty is acceptable? Who records the decision? If the team cannot answer those questions in the exercise, adding another policy document will not settle them.
Record the decision, the responsible owner and the evidence still missing. The decision-record example shows a lightweight way to do that. Where a claim concerns system behaviour, connect it to evaluation evidence. These are proposed product practices, not a finding that a system complies with the law.
Primary timeline reference: European Commission, AI Act. The operational recommendations above are my analysis.
