Health.AI News
Despite requests from national and European industry to postpone the timeline for the introduction of the AI Act for various reasons, the EU Commission has not complied. For deployers, this means that in four weeks the corresponding governance obligations for legally compliant AI use will take effect.
The pending “Code of Practice” for so-called general-purpose models (General Purpose AI — GPAI), such as AI models from OpenAI, Meta, and Google, is now to be presented by the Commission in the near future, so that this part of the EU AI Regulation can also enter into force in August. The Code of Practice is intended to define application rules for GPAI models as best practice for their safe operation within the meaning of the EU Artificial Intelligence Regulation (AI Act). Application of the Code is voluntary, meaning that compliance with the AI Act can also be achieved by other means.
Overview of entry into force and implementation stages of the AI Act:
In force since 1 August 2024
Phased full applicability — Art. 113:
– 2 February 2025: Prohibition of AI systems with unacceptable risk (Art. 1–5)
– 2 August 2025: In particular, sanctions (Art. 99 ff.) and codes of conduct and rules for general-purpose AI (Art. 51 ff.)
– 2 August 2026: General applicability of most provisions, except Art. 6(1) and the corresponding obligations
– 2 August 2027: Obligations for high-risk AI systems (Art. 6(1))
Dr. Britta Rosenberg, Legal Expert in the HEALTH.AI Pomerania project, University of Greifswald, Institute of Pharmacy / Prof. Dr. W. Weitschies