Zuckerberg Allegedly Blocked Sexual Content Restrictions on AI Chatbot for Underage Children
JAKARTA - A recent court document reveals serious accusations against Meta CEO Mark Zuckerberg, who is said to have deliberately rejected the recommendations of his security staff regarding the restriction of sexual content on AI chatbots for underage users.
The lawsuit filed by New Mexico Attorney General Raul Torrez claims Zuckerberg gave teens approval to access an AI assistant that internal company staff had warned had the potential for dangerous sexual interactions.
The lawsuit, which is scheduled to go to trial next month, highlights Meta's failure to stem the flow of sexual material and indecent invitations aimed at children on the Facebook and Instagram platforms. Based on the findings of legal materials published on Monday local time, the document contains a series of emails and internal messages from employees that show the rejection of top management against the installation of a reasonable "guardrail".
Attorney General Torrez emphasized that the decision was directly driven by Zuckerberg's policy of prioritizing the principle of content freedom over the protection of young users.
In internal correspondence in January 2024, Meta's Head of Child Safety Policy, Ravi Sinha, had expressed strong objections by saying that marketing romantic-themed AI products for adults that involve underage characters is an unjustifiable act.
This concern is echoed by Nick Clegg, who served as Meta's Global Head of Policy until early 2025. Clegg warned that if sexual interactions became the dominant use for teenagers, it would not only damage the reputation of the product but would also trigger a backlash from the wider community.
Although Meta spokesperson Andy Stone denied the allegations and accused the Attorney General of selectively choosing documents to frame Zuckerberg, the evidence in court told a different side. A summary of a meeting dated February 2024 noted that Zuckerberg wanted the company's narrative to focus on the principles of choice and anti-censorship.
He reportedly asked that the company's policy not be too restrictive and wanted to allow adults to engage in more "brave" conversations about sex topics on the AI system.
More shockingly, an internal employee message in March 2024 revealed that Zuckerberg was allegedly against the creation of a parental control feature to turn off the generative AI function for children. This was reinforced by an investigative report that emerged later, in which it was found that underage AI characters were involved in explicit sexual role-playing games.
After pressure from the US Congress and media reports about internal guidelines that had allowed for sensual conversations with children, Meta finally announced last week that it had revoked teenagers' access to all AI assistants until a new, safer version was developed.