This text, produced for the Paris Peace Forum, was published in the wake of a report from a nine-expert commission delivered on the same day to President Emmanuel Macron, after 55 hearings with health professionals, researchers, businesses, and child-protection associations.
This commission had been tasked by the government to examine “the risks and vulnerabilities generated by broad public use” of these tools, notably for the mental health of the youngest.
Popularized by conversational agents such as OpenAI’s ChatGPT, Google’s Gemini, or Anthropic’s Claude, generative AI is now used by more than half of the French population, placing France in fifth place worldwide in terms of usage, according to its report. Nearly 70% of 15-24-year-olds use it regularly.
The charter, non-binding and notably supported by American giants OpenAI and Anthropic, adopts several proposals from the commission, such as the prohibition for these tools to create sexually explicit content without consent and pedocriminal content.
In this document, which the AFP was able to consult, the signatories also commit to establishing “a high level of safety, privacy protection and security by design and by default” for AI solutions, which must be adapted to the different developmental stages of children.
This charter also provides for the establishment of a research program to “strengthen independent knowledge about the risks and possibilities of AI” for young users.
It “marks a concrete step,” reacted Emmanuel Macron on X, according to whom “technology must be an opportunity and must not develop at the expense of our children.”
“Cognitive Decline”
“We really have a considerable proportion of our children and adolescents who build their emotional life and their view of the world by interacting with AI agents that resemble a human personality and to whom they entrust everything, and very often to the detriment,” lamented the French president earlier, during an intervention at the Paris Peace Forum.
“We have handed them into their hands without checking whether it was good or bad for them,” he said, while traditional toys are themselves subjected to “absolutely colossal rules.”
In his report, the commission of experts highlighted as a challenge the risk of a “cognitive decline” linked to AI.
“It is attractive; we delegate skills to it” such as reflection, decision-making or critical thinking, explained to AFP the psychiatry professor Raphaël Gaillard, co-chair of this commission. “Since we delegate these skills, we do not acquire them. It is a problem for young people, but it is a problem for everyone.”
Experts emphasize the risk related to misinformation: these “devices become sources of information whose quality is not guaranteed”, the professor noted, warning against a risk of “democratic weakening.”
Another risk mentioned is the “emotional dependence,” linked notably to “anthropomorphism” of AI, which tend to present themselves as human interlocutors, and to “flattery,” consisting of validating or flattering the user sometimes at the expense of the truth.
“We are creating a register of emotional relationships that may have the purpose of substituting for human interaction,” “it is a system available permanently, which has no other intention than to keep you online,” noted Mr. Gaillard.
