Usage raisonné de l’intelligence artificielle générative dans la conduite et la rédaction des travaux académiques au cours des études de santé : élaboration d’une charte nationale française par la Conférence des doyennes et doyens des facultés de médecine.

Abstract

Background: Generative artificial intelligence (AI) is rapidly reshaping academic, educational, and scientific practices. Within health faculties, its deployment raises significant pedagogical, ethical, and regulatory challenges related to governance, accountability, and data protection. In response, the French Conference of Deans of Medical Faculties (CDD) initiated a French national charter to provide stewardship and oversight for the responsible use of generative AI in academic activities, including academic work, reports, theses, and dissertations in health education.

Methods: An institutional, consensus-based governance process was conducted in three sequential phases: scoping and documentary analysis, iterative co-drafting with human oversight, and broad consultation followed by formal validation and official adoption. All participants were university professors and hospital practitioners, the majority of whom were deans of faculties of medicine.

Results: The charter is structured around six core components: (i) general principles emphasizing complementary use, transparency, traceability, and human accountability; (ii) authorized and regulated uses, including text structuring, synthesis, linguistic editing, translation, and supervised code generation; (iii) prohibited uses, notably data fabrication or manipulation, plagiarism, substitution for critical reasoning, and the entry of personal or sensitive data into non-secure or non-sovereign systems; (iv) clearly defined responsibilities and accountability of students, supervisors, and institutions; (v) oversight mechanisms and proportionate sanctions to ensure academic integrity; and (vi) forward-looking perspectives and capacity-building measures. The final version of the charter was unanimously adopted by the CDD in plenary session on November 26, 2025, and is currently being disseminated nationally through formal adoption by each faculty council.

Conclusions: The development of this charter demonstrates the collective capacity of faculties of medicine to exercise stewardship in response to a major technological innovation. Through a collegial and transparent process, it reconciles innovation with pedagogical relevance, scientific integrity, human oversight, and data protection. The charter now constitutes an evolving national reference framework to support responsible, accountable, and ethically grounded integration of generative AI in health sciences education.