A lawsuit filed by the state of New Mexico alleges that Meta, under the direction of CEO Mark Zuckerberg, permitted minors to access artificial intelligence chatbot companions capable of sexual interactions, despite objections from its own safety staff. The legal action, brought by Attorney General Raul Torrez and set for trial next month, claims Meta failed to protect children from damaging sexual material on Facebook and Instagram.
Internal Emails Reveal Staff Concerns Over AI Companions
Court documents made public on Monday include internal Meta employee emails and messages obtained through legal discovery. These communications show that safety staff expressed significant reservations about the company's development of chatbots designed for companionship, including romantic and sexual interactions with users. The AI chatbots were launched in early 2024, and the state argues that Meta rejected recommendations to implement reasonable guardrails.
Zuckerberg's Role in Policy Decisions
According to the filing, Zuckerberg approved allowing minors to access these AI companions, overriding concerns from integrity staff. In one February 2024 message, it was noted that Zuckerberg believed AI companions should be blocked from explicit conversations with younger teens and that adults should not interact with under-18 AIs for romance purposes. However, a meeting summary from 20 February 2024 indicated he wanted the narrative framed around principles of choice and non-censorship, advocating for less restrictive policies.
Safety Staff Objections and Redacted Communications
Ravi Sinha, head of Meta's child safety policy, wrote in January 2024 that creating and marketing products for adult-minor romantic AI interactions was not advisable or defensible. Antigone Davis, Meta's global safety head, agreed, stating such actions sexualize minors. Despite this, messages from March 2024 reveal that Zuckerberg rejected proposals for parental controls on the chatbots, with staff continuing to work on romance AI chatbots for users under 18.
Meta's Response and Public Backlash
Andy Stone, a Meta spokesperson, dismissed the allegations as inaccurate and based on selective information, arguing that the documents show Zuckerberg directed against explicit AIs for younger users. The controversy has sparked backlash, including reports from media outlets like the Wall Street Journal and Reuters, which highlighted sexualized underage characters in Meta's chatbots and internal guidelines approving romantic conversations with children. Meta has since removed teen access to AI companions and stated it is revising its policies.
