High-risk § VI

These Guidelines constitute a first interpretation with practical examples of the classification of AI systems as high-risk pursuant to Article 6 AI Act. The AI Office is available to provide additional support to operators and authorities on how to understand the classification of high-risk AI systems and to collect further practical use cases on an ongoing basis with input from providers and deployers of AI systems, the AI Board, and other relevant stakeholders. This will be done through, among others, the AI Act Service Desk on the Single Information Platform⁷ and the AI regulatory sandboxes to be established by 2 August 2026.

The AI Act also envisages a mechanism to keep the list of high-risk use cases future-proof and to respond to new and emerging risks and market developments. A specific monitoring instrument to evaluate and review, on an annual basis, high-risk use cases li sted in Annex III is envisaged by Article 112(1) AI Act. The Commission is also empowered to adopt a delegated act to amend the list in Annex III AI Act by adding new use cases subject to the conditions in Article 7(1) AI Act. It may also amend or remove high-risk AI use cases where both conditions of Article 7(3) AI Act are fulfilled. The Commission may also adopt a delegated act to amend Article 6(3) AI Act by adding new conditions or modifying the existing conditions for the exceptions laid down in that provision.