Based on this „European approach“ to AI, this article analyses the current European regulatory framework for medical AI. Based on the fundamental rights framework as clear guidelines, an in-depth review of the different areas of law with a focus on data protection, product approval procedures and liability law is then launched. This analysis of the current legal situation, including its problems and ambiguities regarding AI, is complemented by an overview of proposed changes to product authorisation procedures and liability law which, by advocating a human-centred approach, will fundamentally influence how medical AI and AI in general will be used in Europe in the future. PHG Foundation: Legal Accountability for Machine Learning in Healthcare (2018). www.phgfoundation.org/briefing/legal-liability-machine-learning-in-healthcare Hacker, P., Krestel, R., Grundmann, S., Naumann, F.: Explainable AI under contract and tort law: legal incentives and technical challenges. Artif. Intell. Law 16 (2020). doi.org/10.1007/s10506-020-09260-6 O`Sullivan, S., et al.: Legal, regulatory and ethical frameworks for the development of standards in artificial intelligence (AI) and autonomous robotic surgery. J. Med. Robot.

Calculate. Help. 15, 1–12 (2019). doi.org/10.1002/rcs.1968 At the end of February 2020, the European Commission published a White Paper on artificial intelligence (AI) and an accompanying report on the impact of AI, the Internet of Things (IoT) and robotics on safety and liability. In its White Paper, the Commission highlighted the „European approach“ to AI, stressing that „it is crucial that European AI is based on our values and fundamental rights such as human dignity and privacy“. It also announced its intention to propose EU legislation on high-risk AI applications in the near future, which will cover the majority of medical applications of AI. Cohen, I.G.: Informed consent and medical artificial intelligence: what to say to the patient? Georgetown Law J. (2020). papers.ssrn.com/sol3/papers.cfm?abstract_id=3529576 Harned, Z., Lungren, M.P., Rajpurkar, P.: Machine Vision, Medical AI and Malpractice. Harv. J. Law Technol.

Digest (2019). jolt.law.harvard.edu/digest/machine-vision-medical-ai-and-malpractice PHG Foundation: Algorithms as Medical Devices (2019). www.phgfoundation.org/documents/algorithms-as-medical-devices.pdf FDA: Proposed Regulatory Framework for Changes to Software Based on Artificial Intelligence and Machine Learning (AI/ML) as a Medical Device (SaMD) – Discussion Paper and Request for Comment (2019). www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-software-medical-device Brkan, M.: Do algorithms rule the world? Algorithmic decision-making and data protection under the GDPR and beyond. J. Law Inf. 27, 91-121 (2019). doi.org/10.1093/ijlit/eay017 Cabitza, F., and Zeitoun, J-D. (2019). Pudding Proof: A Praise of a Real-World Validation Culture for Medical Artificial Intelligence. Ann. trad.

Med. 7 (8), 161. doi:10.21037/atm.2019.04.07 Minssen, T., Gerke, S., Aboy, M., Price, N., Cohen, G.: Regulatory responses to medical machine learning. J. Biosci Law. 1–18 (2020). doi.org/10.1093/jlb/lsaa002 European Commission (2019a). Communication on building trust in human-centric artificial intelligence. Brussels, 168.COM Available at: eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52019DC0168&from=BG (accessed 26 October 2021). Schönberger, D.: Artificial Intelligence in Healthcare: A Critical Analysis of Legal and Ethical Implications. 27, 171–203 (2019).

doi.org/10.1093/ijlit/eaz004 Quote: Meszaros J, Minari J and Huys I (2022) The future regulation of artificial intelligence systems in health services and medical research in the European Union. In front of-. Genet. 13:927721. doi: 10.3389/fgene.2022.927721 However, there are exceptions to the above rule, such as the concept of a „regulatory sandbox“. A „regulatory sandbox“ is a „safe space where companies can test innovative products, services, business models and delivery mechanisms without immediately drawing all the normal regulatory consequences of carrying out the activity in question“ (Financial Conduct Authority, 2015). Regulatory sandboxes were first used in the FinTech sector, but have expanded to other sectors, including healthcare (Leckenby et al., 2021; see also Fenwick et al., 2018). The AI Act will provide a legal basis for the processing of personal data for the development of certain AI systems in the public interest under the AI regulatory sandbox, in line with the GDPR.38 The proposal is part of a broader set of measures on AI that also includes the updated coordinated plan for AI. Together, the regulatory framework and the coordinated plan will ensure the safety and fundamental rights of individuals and businesses in relation to AI.

And they will strengthen the adoption, investment and innovation of AI across the EU. European Commission (2021). Communication on promoting a European approach to artificial intelligence. Brussels. Available from: digital-strategy.ec.europa.eu/en/library/communication-fostering-european-approach-artificial-intelligence (accessed March 11, 2021). Edwards, L., Veale, M.: Enslaving the algorithm: from a „right to an explanation“ to a „right to better decisions“? IEEE Secur. 16, 46–54 (2018). doi.org/10.1109/MSP.2018.2701152 The AI Bill lists high-risk AI systems in the eight specific areas listed below.