Regulation (EU) 2024/1689

High-risk employment AI: evidence before deployment

How employment use changes AI Act classification, deployer duties, human oversight, monitoring, and worker information.

Written and reviewed by Fredrik Surtell ·

Which employment AI uses are listed as high-risk?

Annex III covers specified systems used for recruitment, selection, decisions affecting work relationships, task allocation based on personal traits or behaviour, and monitoring or evaluating workers.

Operating sequence

  1. 01

    Document the exact intended purpose and whether an exclusion or classification exception is claimed.

  2. 02

    Assess instructions, input data relevance, human oversight, logs, monitoring, worker information, and impact-assessment duties.

  3. 03

    Coordinate AI Act evidence with GDPR, employment, equality, and collective-rights requirements.

Failure modes to avoid

  • Treating a vendor’s marketing category as the legal classification.
  • Adding a human click without giving the reviewer competence, authority, time, and usable information.
  • Ignoring changes that turn a low-risk feature into a listed employment use.

The source remains the authority.

Regulation (EU) 2024/1689, Annex III — EUR-Lex