FRANCE Trends and Developments Contributed by: Mathilde Merckx, Xavier Lebrasseur, Juliette Migeon and Diane Courtier, Alchimie Avocats
the human eye, optimising treatment respons- es, improving access to healthcare by enabling remote patient monitoring and treatment, etc. The possibilities unlocked by AI in healthcare are vast – but so are the associated risks. Acceler- ating the use of AI for patients’ treatments and care automatically raises the following question: what happens in case a patient is harmed? Who should be held liable when a diagnosis is missed due to an AI recommendation-based decision? Is it the healthcare professional who followed the recommendation without question, the hospital that deployed the AI tool or the manufacturer of the AI-enabled medical device? These questions are pressing in France, where applicable legal framework surrounding liability in life sciences and healthcare is complex, and where a new EU Directive is poised to reshape Products Liability Law in France and other EU members states. Current Legal Framework Applicable to AI-related Liability in France Under Article L.1142-1 of the French Public Health Code (“FPHC”), healthcare professionals and institutions are only liable for the harmful consequences of acts of prevention, diagnosis or care in the event of a fault, except where their liability is incurred due to a defect in a healthcare product. This notion of fault can be difficult to grasp in practice, and it is likely that the acceleration of the use of AI in medical practice raises new, as yet unresolved questions. As early as 2023, the National Physicians Body ( Conseil national de l’ordre des Médecins ) ques- tioned the impact of new technologies in gen- eral, and AI in particular, in medical practice
on the legal interpretation of fault. This was to better understand the stakes in terms of liability incurred by the healthcare professional who is willing to integrate AI into their daily practice, on the one hand, and the healthcare professional who is reluctant to adopt new technologies on the other. In practical terms, while the former could be held liable for placing blind trust in an AI tool, the latter could be found at fault and held liable for failing to use a technology that could have enabled a patient to benefit from earlier diagnosis or more appropriate treatment. The principle of “loss of opportunities” ( perte de chance ) may then apply in such a scenario, potentially marking a shift in the standard of care for practitioners. On the industrial side, when a health product causes damage to a patient, liability for defective products Law is the most common applicable regime (although plaintiffs and civil courts seem to have increasingly resorted to the fault-based liability regime for the past few years, particu- larly where the defectiveness of the products if hard to prove and where negligence is easier to demonstrate). Current French defective product liability legal framework stems from the transposition of the European Directive 85/374/EEC of 25 July 1985. Under existing rules, producers are liable for the damage caused by a defect in their product, whether or not they have a contractual relation- ship with the victim. Under Article 1245-3 of the French Civil Code (“FCC”), a product is defective when it does not provide the safety which a person is entitled to expect, taking all circumstances into account including the presentation of the product, the time when the product was put into circula-
50
CHAMBERS.COM
Powered by FlippingBook