MedTech Europe responds to stakeholder consultation on the classification of high-risk AI systems
Posted on 11.09.2026
The Artificial Intelligence Act (AI Act), published as part of a wider package of policy measures supporting the development of trustworthy AI in the EU, lays out a set a harmonised rules to address the risks of AI and position the Union as a global player. As part of these rules, it distinguishes between two types of “high-risk” AI systems, namely:
- AI systems embedded in products regulated under the Union’s harmonised legislation on product safety.
- AI systems that can significantly affect people’s health, safety, or fundamental rights in specific use cases listed in the AI Act.
To help providers and deployers assess whether their AI systems fall under the “high-risk” classification and to which type, the Commission has developed a set of guidelines. These guidelines were open to feedback from relevant stakeholders until 23 July 2026. With these guidelines, the Commission aims to provide clarity on the relevant provisions of the AI Act, while including practical examples to help providers, deployers and competent market surveillance authorities in their assessment of their AI system.
As the trade association representing the medical technology industry, MedTech Europe highlighted areas where the draft guidelines could create legal uncertainty and stressed the importance of ensuring alignment with the outcome of the Digital Omnibus on AI.
In its contribution, MedTech Europe put a focus on the special nature of medical devices, which often combine rule-based and AI-based functionalities, a particularity which might result in uncertainty as to whether the entire system of only specific components of the device should be considered AI.
Amongst other, the response also calls for further clarification for the definition of an AI system as well as for the distinction between AI systems which are components of products and those which are products in their own right.
As a result from the Digital Omnibus on AI, key AI Act obligations have seen their date of application be postponed. Under the revised timeline, obligations for high‑risk stand‑alone AI systems will apply from 2 December 2027, while AI systems used as safety components and regulated under the MDR and IVDR will become subject to the AI Act from 2 August 2028.
For more information, please reach out to Leander Vranken, Manager Digital Health AI.