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.
Simplification deserves a specific question: which work has become simpler, and which protection remains? A shorter form and a later deadline are different changes. Treating them as synonyms makes it harder to judge either one.
Evaluate the change, not the label
For a product team, the useful unit of analysis is the control. Suppose a reporting process is shortened. Does the team still record the system’s intended use, the relevant risk and the evidence behind the release decision? If yes, the reduction may remove duplication. If the evidence disappears, the change has a different consequence.
This is why I am cautious about both celebratory and catastrophic headlines. They can skip the actual comparison. A change register should show the previous requirement, the current requirement, the affected systems and the implementation decision. That is less exciting than declaring victory, but considerably more useful on Monday morning.
Consider an illustrative team maintaining the same risk description in three documents. Reducing that to one controlled record could improve consistency. Removing the person responsible for reviewing it would be a different decision entirely. The number of pages tells you very little about the quality of oversight.
| Decision | Evidence to request |
|---|---|
| Remove duplication | One authoritative record remains |
| Change a review step | Responsibility and evidence remain clear |
| Extend a deadline | Work plan and interim controls remain visible |
This is not entirely cynical. The EU acknowledged a real sequencing problem. Asking companies to comply with standards that do not yet exist creates a strange kind of regulatory theatre. The co-legislators knew it, so they adjusted. That part makes sense.
Regulators gave you more runway. That is not the same as telling you the risk is not real.
What to do with this
Ask reviewers to distinguish a reduction in administration from a reduction in protection. If those are bundled together, make the trade-off explicit. Otherwise “simplification” becomes a suitcase large enough to hide every decision nobody wants to explain.
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.
