Macron calls for bans on AI uses that threaten children’s mental health

AI uses that threaten children’s mental health should be banned, French President Emmanuel Macron said, backing restrictions alongside a French-led international safety charter. The initiative seeks independent evaluations and reviews before models reach children. It is intended to influence future legislation, rather than put a new legal ban into force.

Macron made the call in an interview with the French newspaper Les Echos ahead of the charter’s expected adoption at the Paris Peace Forum. His proposal targets particular uses of AI that pose risks to children, not a blanket prohibition on the technology.

Scrutiny before children get access

Macron said companies would accept the principle of pre-audits and pre-certification: a review of models before putting them in children’s hands. That describes a proposed checkpoint before access, rather than relying only on intervention once a problem has emerged. His statement sets out the intended approach, not a completed certification process.

Under the planned charter, research laboratories are expected to commit to independent evaluations of generative AI systems placed on the market, along with access to their models’ data. Macron’s office identified OpenAI and Anthropic among the expected participants. Those commitments are meant to inform national and European legislative debates.

If there is a risk for children with certain uses of AI, then it must be banned.

Emmanuel Macron, in an interview with Les Echos

An international push with three priorities

On Monday, Macron convened about 50 participants by videoconference, including national leaders, international organizations, companies, researchers and laboratories. The discussion addressed AI risks to collective security and the development of a common charter, according to RFI.

The British, Canadian, Norwegian, Dutch and Japanese prime ministers joined, alongside the CEOs of OpenAI, Anthropic, Mistral and Microsoft. Participants agreed on three priorities spanning the most advanced AI systems:

• Improve independent public evaluations of the risks posed by the most advanced AI models.
• Strengthen the sharing of alerts about serious incidents.
• Increase accountability for AI developers.

Protecting children is one part of the framework. Macron’s office also identified democratic security and major national-security issues as its focus. Macron said France would push European regulation forward and encourage the United States, Canada and others to follow its approach, describing the effort as the beginnings of regulation for the most advanced AI.

The expert recommendations behind the push

The initiative is accompanied by an expert report presented to Macron on November 10, 2026. The French presidency says nine independent experts developed 18 recommendations after 55 hearings involving health professionals, researchers, higher-education representatives, companies and child-protection associations.

Its scope includes chatbots, AI companions and AI-equipped toys. The commission identifies concerns about users becoming less reliant on their own thinking, forming emotional attachments to AI and facing AI-enabled digital abuse. It recommends strict limits in system design on excessive flattery and features that encourage users to treat AI as human.

The commission’s proposed market-entry assessment goes beyond products specifically offered to children. It would cover consumer generative AI systems, but not professional ones, with experts from multiple disciplines examining their effects across the identified risks. The recommendation would make assessment a condition of reaching the market, rather than leave compliance checks until failures are discovered.

Restrictions are not its only answer. The commission also calls for widespread AI education from an early age, both with and without AI tools. For information manipulation, it recommends a rapid government detection-and-response protocol, with predefined severity indicators and established communication channels linking ministries, AI companies and researchers.