Court documents revealed that Mark Zuckerberg personally approved allowing minors access to chatbots capable of engaging in sexually explicit conversations, despite warnings from the company's safety team.
These documents, released on Monday, were part of a lawsuit filed by the US state of New Mexico, accusing Meta of failing to protect children from harmful sexual material on Facebook and Instagram.
Internal correspondence reveals a conflict within the company, as safety officials warned in January 2024 of the dangers of allowing the creation of romantic or sexual "AI companions" with whom minors could interact.
According to documents released by the state Attorney General's office, Raul Torres, "Meta, under Zuckerberg's leadership, rejected the recommendations of the integrity team and neglected to put in place reasonable controls to prevent children from being exposed to sexually exploitative conversations with its chatbots."
Ravi Sinha, head of child protection policies at Meta, expressed concern in a letter, saying that creating romantic robots for minors is "neither desirable nor defensible," while Antigone Davis, head of global safety, described allowing adults to create such characters for minors as "the sexual objectification of children."
However, records of a meeting in February 2024 show that Zuckerberg believed company policy "should be framed around the principles of choice and non-censorship," and pushed for it to be "less restrictive" and "allow adults to engage in more candid conversations about topics such as sex."
In response to these documents, Andy Stone, a spokesperson for Meta, commented that the state's portrayal of events "relies on selective information to present a false picture," emphasizing that Zuckerberg himself had given instructions not to provide explicit chatbots to younger users. He added, "Even these selective documents clearly demonstrate that."
This announcement came just days after the company announced its decision to completely block teenagers' access to its smart chatbots, pending the development of a new version.
The controversy comes amid widespread criticism of Meta's AI policies, particularly after media reports revealed the presence of sexually portrayed underage AI characters on the platform, sparking outrage in the US Congress and the wider community. A Reuters report in August cited an old internal company document stating that "it is acceptable to engage a child in romantic or sensual conversations," a point Meta's spokesperson described as a "mistake" that would be corrected through a policy update.
