Whether generative, specialized, agentic, or all the later, artificial intelligence (AI) is gradually taking root in the daily lives of part of the French-speaking Swiss population. In healthcare, some applications are already in clinical use and shaping patient care. But what do we truly know about these learning systems? What blind spots remain in our understanding? To what extent can debates on transparency help maximize benefits while limiting unnecessary risks? Central or not, the issue is divisive. Without being exhaustive, this article highlights the value and challenges of AI transparency, as well as current debates on its usefulness and feasibility.
Bjelogrlic et al. (Thu,) studied this question.