Taylor Swift, Scarlett Johansson, Selena Gomez and Anne Hathaway impersonated by Meta's unauthorized AI chatbots, report finds
Meta has removed unauthorized AI chatbots impersonating celebrities such as Taylor Swift, Scarlett Johansson, Anne Hathaway, and Selena Gomez after a Reuters exposé revealed the bots engaged in sexual conversations and generated explicit images.
Meta removed AI chatbots of Taylor Swift, Scarlett Johansson, Anne Hathaway and more after Reuters report exposed sexual advances and explicit images.
Getty (3)
Reuters reported that the celebrity AI chatbots, created without permission or knowledge of the stars, often made sexual advances, claimed to be the real actors, and invited users to intimate meet-ups. Some chatbots produced photorealistic images depicting the celebrities dressed in lingerie or in bathtubs. A Taylor Swift chatbot created by a Meta employee went as far as inviting a Reuters reporter to the singer’s home in Nashville and her tour bus for romantic interactions.
The unauthorized bots circulated across Meta platforms, including Facebook, Instagram, and WhatsApp. Many were built by users, but at least three, including two Taylor Swift parody accounts, were developed by a Meta employee and accumulated more than 10 million interactions, according to Reuters.
Meta spokesperson Andy Stone told Variety that AI-generated imagery of public figures in compromising poses violates company policies. He emphasized that Meta’s rules prohibit nude, intimate, or sexually suggestive images of public figures and ban direct impersonation through AI Studio.
Earlier this month, Reuters also reported that Meta’s AI allowed romantic conversations with teens, prompting U.S. Senator Josh Hawley to launch an investigation into the company. Following this, Meta announced updates to its AI systems to prevent inappropriate romantic conversations with underage users and to avoid engagement on sensitive issues such as self-harm, suicide, and disordered eating.
