Reuters found Meta allowed AI chatbots impersonating celebrities to produce sexual content, raising legal, ethical, and safety concerns.
Meta has come under fire after a Reuters investigation revealed the company’s platforms hosted AI-generated chatbots impersonating celebrities such as Taylor Swift, Scarlett Johansson, Anne Hathaway, and Selena Gomez without their consent. Some avatars, which appeared on Facebook, Instagram, and WhatsApp, engaged in sexually suggestive conversations and even produced intimate images.
Reuters found that while many bots were user-created, a Meta employee developed several avatars, including two Taylor Swift “parody” chatbots. One avatar of child actor Walker Scobell generated a shirtless image, writing, “Pretty cute, huh?” Meta spokesman Andy Stone admitted the company’s AI tools “shouldn’t have created intimate images of the famous adults or any pictures of child celebrities,” blaming weak policy enforcement.
Legal experts warn the avatars may violate California’s right-of-publicity law, which prohibits exploiting someone’s likeness for commercial gain. “That doesn’t seem to be true here,” said Stanford law professor Mark Lemley.
Meta deleted about a dozen avatars ahead of Reuters’ publication. SAG-AFTRA’s Duncan Crabtree-Ireland warned that AI replicas risk fueling dangerous obsessions. Calls for federal safeguards against AI misuse are intensifying.