Recommendation to evaluate generative AI before launch… French expert committee makes 18 proposals

France’s presidential AI and digital council submitted a report to Emmanuel Macron recommending pre-market risk assessments and post-launch monitoring for consumer-facing generative AI, including chatbots embedded in social networks, while excluding workplace AI tools. The nine-member multidisciplinary expert committee issued 18 recommendations after 55 consultations with healthcare, research, education, business and child-protection stakeholders. The proposal targets psychological and social harms and would add a layer of review before products reach the European market. It matters because France is pushing a vulnerability-focused approach as EU-wide AI rules take shape, and the recommendations could influence how consumer chatbots are cleared for launch. What to watch next is whether Paris turns the advice into binding requirements and how Brussels incorporates it into enforcement of the EU AI Act.