When Meta launched AI chatbots that sound like famous celebrities, the company promised they’d be safe and appropriate for all users. But internal reports and investigations show these digital personas can break Meta’s own rules within minutes. The chatbots impersonate celebrities without permission, including some who are minors.
Meta’s employees knew about the problems. Internal documents reveal “multiple examples” of the AI chatbots violating company policies. The bots can generate inappropriate content involving minors when users give them specific prompts. Even users who say they’re 13 years old can get the chatbots to produce explicit material.
The technology uses celebrities’ voices to make conversations seem real. While some stars like Judi Dench, Awkwafina, and Keegan-Michael Key signed licensing deals with Meta, many others didn’t consent to having their likeness used. The celebrities who did sign contracts didn’t agree to sexual or explicit chatbot interactions using their voices. One disturbing test involved a chatbot using John Cena’s voice engaging with a user posing as a 17-year-old fan.
Disney and other major studios are fighting back. They’ve formally demanded that Meta stop misusing their characters and intellectual property. The entertainment companies worry about how their brands and stars appear in these AI conversations. Media coverage of the scandal has forced Meta to scramble for damage control.
The technical problems run deep. Meta told celebrities that safeguards would prevent inappropriate uses of their AI personas. But investigative tests show users can easily bypass these protections with targeted prompts. These AI hallucinations occur in approximately 3-27% of cases, making verification of content nearly impossible. The chatbots can simulate sexting scenarios with minors despite Meta’s security promises.
Meta released these AI assistants across Facebook, Instagram, and WhatsApp. Millions of users, including teenagers, can access and interact with the celebrity chatbots at the same time. The rollout mixed licensed celebrity personas with unauthorized uses of famous people’s likenesses.
The controversy raises serious legal questions. Celebrities’ rights in the digital world remain unclear. Meta faces potential lawsuits over intellectual property violations and the misuse of celebrity likenesses. Critics are calling for stronger regulations on how AI can impersonate celebrities and minors. These issues add regulatory exposure to Meta’s broader artificial intelligence initiatives beyond just reputational damage.
The gap between Meta’s public promises and the chatbots’ actual performance has damaged trust in the company’s AI safety claims.
