The European Commission is conducting a targeted consultation on draft guidelines for classifying AI systems as high-risk under Article 6 of the EU AI Act (Regulation (EU) 2024/1689). The consultation deadline was extended to 23 July 2026 in response to stakeholder requests, from an original closing date of 23 June 2026. The Commission plans to adopt the final guidelines by end 2026.
Article 6 AI Act sets the classification criteria for high-risk AI systems, with Annex III listing eight specific categories: biometric identification; critical infrastructure management; education and vocational training; employment and worker management; access to essential private and public services; law enforcement; migration and border control; and administration of justice. Article 7 permits the Commission to expand Annex III by delegated act. The draft guidelines interpret the operative concepts within Article 6 and Article 7, supply worked examples of systems that should and should not be classified as high-risk, and set out how Annex III categories map to specific product types.
AI system providers, deployers, importers, and distributors across all sectors should assess their AI portfolios against the draft classification criteria before the guidelines are finalised. A high-risk classification triggers mandatory obligations under Title III of the AI Act: conformity assessment under Article 43, technical documentation under Article 11, registration in the EU database under Article 71, post-market monitoring under Article 72, and incident reporting under Article 73. Technology companies, banks, insurers, healthcare providers, recruitment platforms, and public-sector AI deployers in EU member states are the primary categories affected by Annex III.
The guidelines are interpretive, not binding law. The Commission retains discretion to modify the final text in response to consultation submissions. Market surveillance authorities are not formally bound by the guidelines but are expected to give them significant weight in enforcement decisions. Organisations that contest the Commission's proposed classification of their systems have until 23 July 2026 to submit comments through the digital-strategy.ec.europa.eu consultation portal.
Licentium advises on EU AI Act classification strategy and conformity assessment readiness. We may be able to assist clients in preparing substantive consultation responses or assessing their AI portfolios against the draft high-risk criteria, and can refer specialist regulatory counsel through our partner network where needed. Work we undertake includes AI system risk classification, Article 6 scoping analysis, Annex III mapping, conformity assessment preparation, and regulatory consultation drafting.