EU Considers Major Reforms to Medical AI Regulation for Simplification

The European Union is actively considering two distinct yet interlinked regulatory proposals aimed at simplifying and harmonizing the governance of artificial intelligence within medical devices. This initiative stems from widespread industry criticism regarding the current regulatory overlap, where AI medical devices are typically subject to both the Medical Devices Regulation (MDR)/In Vitro Diagnostic Regulation (IVDR) and the EU AI Act's 'high-risk' requirements. This dual compliance framework has been identified as duplicative and burdensome.
The first proposal, termed the 'Digital Omnibus' and spearheaded by DG CONNECT, seeks to streamline the interaction between the AI Act and the MDR/IVDR. Its primary objective is to reduce bottlenecks and eliminate redundant requirements, making it easier for 'notified bodies' to conduct a single, comprehensive assessment for both sets of regulations. This proposal also aims to prevent delays by potentially postponing the application of certain AI Act obligations until necessary harmonized standards are fully established, thereby making 'dual compliance' more practical.
The second proposal, originating from DG SANTE in December 2025, involves a broader amendment to the MDR and IVDR. This more sweeping change suggests that AI medical devices would no longer automatically fall under the high-risk classification of the AI Act. Instead, they would primarily be governed by the MDR/IVDR. Crucially, this proposal includes a provision for the EU Commission to retain the authority to reintroduce specific AI Act requirements for medical AI through future delegated or implementing acts, if deemed necessary.
These diverging paths to simplification highlight the EU's challenge in balancing the promotion of innovation in health tech with the imperative of ensuring patient safety and ethical AI deployment. While both proposals aim to alleviate regulatory burden, the second proposal represents a more significant departure from the current default high-risk classification for medical AI, potentially offering greater flexibility but also raising questions about the long-term consistency and robustness of AI-specific oversight in healthcare. The outcome of these deliberations will profoundly shape the future landscape for AI medical device development and market access within the EU.