Artificial intelligence and medical responsibility1
Press release from the French National Academy of Medicine
December 3, 2025
Developments in artificial intelligence (AI) in the field of healthcare (self-learning systems, generative AI, large language models based on conversational tools, etc.) will lead to significant growth in its use, driven by young practitioners who are mastered in digital resources.
The role of AI in administrative and office tasks, diagnosis, treatment, patient follow-up, prevention, public health, training, research, etc. will only increase with ever more powerful tools. This development is governed by French legislation (Article L4001-3 of the Public Health Code) (1) and by European regulations. These texts do not call into question the French legal principles of medical liability (judicial, administrative, professional), but they do supplement them: emphasis on patient information; introduction of the obligation for doctors to validate data produced by AI (human guarantee); and introduction of shared liability with AI providers, who are also required to guarantee the improvement of their products. In addition, there are obligations to ensure data protection, to allow access to data and, where necessary for its correction (2-5).
Recommended because it is available, the non-use of AI may be ethically reprehensible (6), as it is considered a fault due to loss of opportunity for the patient. However, the doctor is not obliged to strictly follow the AI’s suggestions if they do not seem appropriate.
In light of emerging litigation regarding medical liability related to the use of AI, the National Academy of Medicine recommends that (7):
1- Initial and continuing training for doctors includes a section on the principles governing AI in medicine, the tools used for this purpose, and a critical analysis of its results. This training shall be shared, taking into account skill levels, with the healthcare professionals concerned, patient partners, and patient associations.
2- The doctor informs the patient, in a clear and relevant manner, when an AI system is used for their care, obtain their consent and keep a record of it.
———————————————————————————————–
3- The use of AI does not oblige the doctor to follow its recommendations; the doctor must inform the patient of the reasons for their choice and verify that professional insurance contracts cover the use of AI.
4- Recommendations and p be implemented to support the justification of decisions made and to defend them in the event of a dispute.
5- Serious adverse events related to AI be reported according to a specific, simple and centralized vigilance procedure.
6- Doctors be attentive to the protection of the patient data they import and manage, as the development of AI use makes the conditions for hosting health data and the traceability of patient consent and information regarding this use crucial.
References
– Law No. 2021-1017 of 2 August 2021 on bioethics. JORF No. 0178 of 3 August 2021, 9-41
– European Union. European Regulation 2016/679 of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation). Official Journal of the EU of 4 May 2016, L 119/1-L 119/88.
– Law No. 78-17 of 6 January 1978 on information technology, files and freedoms. JORF No. 0006 of 7 January 1978, 227-231.
– European Union. Regulation (EU) 2024/1689 on artificial intelligence of 13 June 2024 laying down harmonised rules concerning artificial intelligence and amending several regulations. Official Journal of the EU of 12 July 2024, series L, 144 p.
– European Union. Directive (EU) 2024/2853 of 23 October 2024 on liability for defective products and repealing Council Directive 85/374/EEC. Official Journal of the EU of 18 November 2024, series L, 22p.
– Nordlinger B., Kirchner C., de Fresnoye O., Generative AI systems in healthcare: challenges and prospects. Report 24-03 of the French National Academy of Medicine adopted on 5 March 2024, Bull Acad Ntle Med 2024, 308, 536-47.
– Lecomte D., Roques-Latrille C.F.. Artificial intelligence and medical responsibility. What are the challenges? Report by the French National Academy of Medicine adopted on 7 October 2025. https://www.academie-medecine.fr/intelligence-artificielle-et-responsabilite-medicale-quels-enjeux/
1 Press release from the Academy’s Rapid Communication Platform
Bull Acad Natl Med 2025;209:pp-pp. [En ligne] Disponible sur : URL
