Source: IAPP

After various delays, the European Commission released draft guidelines on 19 May and opened a public consultation aimed at helping "providers, deployers and other relevant actors in determining whether an AI system falls within the high‑risk category." The guidance is structured as a three‑phased guide offering general principles for classifying high‑risk AI and providing examples to illustrate how classification should be assessed across different areas and use cases. The draft explicitly addresses the two high‑risk categories under Article 6 of the AI Act: Article 6(1) together with Annex I for AI in the context of product safety, and Article 6(2) together with Annex III for systems deployed across eight sectoral areas including biometrics, education, employment and law enforcement.

The comment period on the draft guidance is open through 23 June, and the Commission does not specify when the guidance will be finalized or whether further public comment rounds will follow. The text notes the guidelines were initially due on 2 February and that the Commission postponed publication to accommodate stakeholder feedback; that delay was linked to the accelerated adoption of the Digital Omnibus on AI (introduced November 2025). Stakeholders — including more than 110 EU‑based businesses — lobbied for a two‑year pause on enforcing high‑risk rules. The provisional agreement on AI Act reforms set new compliance dates: stand‑alone high‑risk systems by 2 Dec. 2027 and high‑risk systems embedded in products by 2 Aug. 2028. The dossier also records a 26 Jan. European Parliament LIBE hearing where Deputy Director‑General Renate Nikolay said the omnibus would allow more time to develop guidelines, specifications or standards.