What happened
The European Commission published draft guidelines on 19 May 2026 (within the broader coverage window given the open consultation running to 23 June 2026) clarifying when AI systems qualify as 'high-risk' under Article 6 of the EU AI Act. The guidelines interpret the two classification tracks (Article 6(1) safety-component products; Article 6(2) Annex III use cases) and provide a non-exhaustive list of examples covering biometrics, critical infrastructure, employment, education, credit scoring, and law enforcement. Stakeholder consultation closed 23 June 2026. The AI Act Advisory Forum held its inaugural session on 19 June 2026 and provided first input on the draft. Final guidelines are expected before August 2026 high-risk obligations take effect.
Why it matters
These guidelines are the authoritative Commission interpretation of the high-risk classification framework — the gateway that determines which AI systems face the heaviest EU AI Act obligations (conformity assessment, technical documentation, human oversight, registration). Broad-purpose and general-purpose AI systems are explicitly addressed. The consultation closure in the window and the Advisory Forum's inaugural review of this draft make it a live, directly actionable development for any organisation deploying AI in the EU.
Action needed
Conduct preliminary AI system classification assessments against the Article 6 criteria using the draft guidelines' examples. Track finalisation closely — final guidelines will bind market surveillance authority interpretation. Consider participating in remaining consultation channels.