"If there is a risk for children with certain uses of AI, then it must be banned. » In a daily interviewThe EchoesPublished on Monday, November 9, 2026 in the evening, and taken up by the AFP, Emmanuel Macron felt that artificial intelligence should be banned where it threatens the "mental health" and "brain" of children. "It is not because it is AI that it must be the Far West," added the Head of State.
A charter with the labs this Tuesday
These declarations take place on the eve of the adoption of acharteat the Paris Peace Forum. According to the Elysée, cited by the AFP, research laboratories, including OpenAI and Anthropic, are required to undertake several measures, such as independent evaluations of generative AI systems put on the market and access to data from their models. According to the president, companies would accept "the principle of pre-auditing and pre-certification" of models before putting them in the hands of children. The charter must "enlighten national and European legislative debates".
An expert committee is also expected to provide the Chair with recommendations on the protection of children from AI on Tuesday, November 10.
Three commitments on advanced model risks
On 9 November, Emmanuel Macron also brought together, in Paris and via videoconference, some 50 heads of state and government, laboratory leaders, researchers and experts on the risks of AI to collective security. According to its message published on X, participants committed themselves to three priorities: to better assess, independently and publicly, the risks of the most advanced models; better share alerts in case of serious incidents; To empower actors who develop models. The participants included the leaders of OpenAI, Anthropic and Mistral AI, as well as the President of Microsoft.
What to keep in mind
At this point, these are announcements and voluntary commitments, not legislation: the exact content of the charter and the committee's recommendations was not public at the time we wrote. The applicable legal framework remains the EU's AI Act, whose next deadline falls on 2 December 2026. For both parents and companies that design services used by minors, the subject of age verification and the "child-friendly" design of assistants may impose rapidement in the debate.