High-Risk AI Systems
On June 6 2025, the European Commission ("Commission") has launched a public consultation aimed at gathering stakeholder feedback on the classification and regulatory treatment of high-risk artificial intelligence ("AI") systems under the EU AI Act, which entered into force on 1 August 2024. This targeted consultation will support the preparation of Commission guidelines that clarify how high-risk AI systems should be identified and subsequently governed, and is posed to play a pivotal role in shaping the implementation of the AI Act's rules before the respective obligations become applicable on 2 August 2026.
High-risk AI systems are subject to stringent legal obligations. However, many questions remain lingering with regard to how these systems should be classified and what pragmatic compliance entails, prompting the Commission to seek stakeholder input at this early stage.
At the time of writing, the AI Act defines two categories of high-risk AI systems under Article 6:
- Those integrated into products subject to EU harmonised legislation on product safety (Article 6(1)); and
- Those deployed in use cases listed under Annex III, which pose significant risks to health, safety, or fundamental rights (Article 6(2)).
The consultation seeks practical examples and input from a wide spectrum of stakeholders to clarify both classification thresholds and obligations throughout the AI value chain. Additionally, the consultation also invites comments on the exemptions mentioned under Article 6(3), and the potential overlaps with prohibited practices under Article 5.
As for the specific guidelines concerned, Article 6(5) paired with Article 96(1)(a) upholds that the Commission shall issue two sets of guidance by 2 February 2026, which include Guidelines clarifying high-risk classification criteria, illustrating detailed practical examples of what comprises of a high-risk system; and Guidelines on fulfilling requirements and obligations applicable to such high-risk systems, including responsibilities along the AI value chain.
The consultation questionnaire covers five key areas, and stakeholders are free to participate in any of the following sections:
- Classification under Product Safety Legislation (Article 6(1) and Annex I)
- Use Case-Based Classification (Article 6(2) and Annex III)
- Cross-Cutting Classification Issues
- Obligations and Governance Across the Value Chain
- Potential Amendments to Risk and Prohibition Lists
The targeted consultation will be open until 18 July 2025. For more information and submissions, kindly access the EUSurvey portal.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
