Agent Governance Register

Reviewed

EU AI Act crosswalk

Regulation (EU) 2024/1689, the Artificial Intelligence Act

In short: The EU AI Act sets duties for high-risk AI systems, including human oversight, record-keeping and log retention. Five register controls map to articles we read on EUR-Lex. Whether a given agent is high-risk depends on its use, so check scope first. This is not legal advice.

This is not legal advice.

What does EU AI Act cover?

These duties apply to high-risk AI systems as the Act defines them. An agent is not high-risk by default; it depends on its use case.

Which register controls map to EU AI Act?

ControlReferenceWhat the text says
GOV-02 Named accountable ownerArticle 26(2)Deployers shall assign human oversight to natural persons who have the necessary competence, training and authority.
GOV-07 Human approval for high-risk actionsArticle 14(4)(d)People assigned oversight must be able to decide not to use the system, or to disregard, override or reverse its output.
GOV-11 Revocation and kill switchArticle 14(4)(e)People assigned oversight must be able to interrupt the system through a stop button or similar procedure that halts it in a safe state.
GOV-13 Per-action audit log tied to a personArticle 12Record-keeping: high-risk systems must allow events to be recorded automatically as logs.
GOV-23 Evidence retentionArticle 19(1) and Article 26(6)Automatically generated logs must be kept for a period appropriate to the intended purpose, of at least six months.

What did we leave out, and why?

We did not map testing (GOV-09) to Article 9 or purpose (GOV-03) to the instructions for use. We did not confirm the exact paragraph text for either.

If you can point to the official text that supports a mapping, write to the desk.

Reviewed by The Register desk