Analysis ·
Managing AI risks: a vital urgency
Original title, in French: “Encadrer les risques liés à l’IA, une urgence vitale”
- Publication
- IT for Business (France)
- Author
- Patrick Brébion
- Published on
- 13 October 2026 · read in the page structured data
Summary
The article, part of a special report on IT and AI in hospitals, describes the risks of using AI in medicine: hallucinations, unsuitable training data, diluted responsibility for doctors, and loss of skills. Jérémie Thomas, product manager at Ethik-IA, explains that his company supports institutions such as Hôpital Foch, CHU de Toulouse and CH d'Avignon in putting an ethical framework around these uses and anticipating legal risks, particularly with regard to the AI Act. This support includes checking the quality of training data. Patrick Callier, a genetics practitioner at CHU de Dijon, illustrates this risk: in dermatology, an AI trained on white skin may give aberrant results on olive skin. The article is truncated for subscribers.
Summary based on the full text of the article.
Read the article on IT for Business →Classification
- Country
- France
- Company size
- All sizes
- Industries
- Healthcare, pharmaceuticals, social care
- Companies
- CH d'Avignon (deploys AI), CHU de Dijon (cited), Ethik-IA (supplier), Hôpital Foch (deploys AI), CHU de Toulouse (deploys AI)
- Keywords
- medical AI, ethical framework, ai act, algorithmic bias, hospital, training data
Entry written by Marsanne Studio from the cited article, with the help of an AI model (claude-sonnet-5). The rights to the article and its illustrations belong to IT for Business and its author. Entry translated from French by Claude Sonnet 5.5.