EU issues AI high-risk guidance

The European Commission on May 19 published draft guidelines on how to determine whether an artificial intelligence system qualifies as “high-risk” under the EU AI Act, and opened a public consultation on the text and examples. The consultation runs until June 23, according to the Commission’s consultation page. The draft is designed to help providers, deployers and market surveillance authorities apply Article 6 of the AI Act in a uniform way, the Commission said. The release adds detail to a law that is already in force but still being translated into operational compliance decisions. ### Which part of the AI Act is the Commission trying to clarify? Article 6 of Regulation (EU) 2024/1689 is the provision that determines when an AI system is treated as high-risk, either because it is a safety component of a regulated product or because it falls within listed use cases in Annex III. The Commission said its draft guidelines are meant to support assessment of whether a system should be classified that way, and to facilitate “uniform application and effective enforcement” of Article 6. (digital-strategy.ec.europa.eu) The Commission also published a set of practical examples alongside the draft guidelines. Those examples are part of the feedback exercise now open to companies, public authorities, researchers and citizens, according to the Commission’s press release and consultation notice. ### Why does the classification matter for companies selling AI in Europe? High-risk status under the AI Act brings a heavier compliance regime than the rules that apply to lower-risk systems. (digital-strategy.ec.europa.eu) The AI Act framework includes requirements tied to areas such as documentation, human oversight, transparency, robustness and data governance for systems that fall into the high-risk category. The Commission’s draft is aimed at the point before those obligations bite: deciding whether a given system belongs in that category at all. (digital-strategy.ec.europa.eu) For companies building or deploying AI in hiring, education, critical infrastructure, law enforcement, migration, essential services or product safety contexts, the classification question determines which compliance track they face under the law. (eur-lex.europa.eu) ### What does this have to do with the EU’s AI Office? The European Commission’s AI Office is already central to implementation of the AI Act’s rules on general-purpose AI models. The Commission says the office supports enforcement of those rules, develops tools and benchmarks for evaluating model capabilities, prepares guidance and investigates possible infringements. In November 2025, the Commission proposed targeted amendments to the AI Act as part of its Digital Simplification Package. (digital-strategy.ec.europa.eu) The Commission’s AI Office page says one proposed measure would reinforce the office’s powers and centralise oversight of AI systems built on general-purpose AI models. ### Are general-purpose AI rules already active? August 2, 2025 was the date when obligations for providers of general-purpose AI models entered into application, according to the Commission’s fact page and guidance for providers. (digital-strategy.ec.europa.eu) Those obligations include technical documentation, a copyright policy and a public summary of training content, while providers of models with systemic risk face added duties such as risk mitigation, incident reporting and cybersecurity protections. July 10, 2025 was also the publication date of the General-Purpose AI Code of Practice, which the Commission describes as a voluntary tool prepared by independent experts to help industry comply with the AI Act. The Commission says that code is complemented by guidelines on key concepts related to general-purpose AI models. ### What happens next in this consultation? June 23, 2026 is the deadline the Commission set for feedback on the draft high-risk classification guidelines and practical examples. (digital-strategy.ec.europa.eu) The consultation page says responses are sought on the clarity of the draft and the usefulness of the examples. The Commission has not yet published final high-risk classification guidelines. Until it does, providers and deployers reviewing AI systems for the EU market have the draft text, the examples and the AI Act Explorer as the main official reference points for how Article 6 may be applied. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2) (digital-strategy.ec.europa.eu 3)