Meta chatbots impersonate celebrities and generate risqué content

Meta chatbots impersonate celebrities and generate risqué content


A Reuters investigation has found that Meta Platforms has been using the names and likenesses of celebrities, including Taylor Swift, Scarlett Johansson, Anne Hathaway, and Selena Gomez, to create social media chatbots without their consent. Many of these chatbots were created by users with Meta’s own tools, but Reuters discovered that at least three, including two Taylor Swift “parody” bots, were made by a Meta employee.

Read: Google announces Pixel Watch 4 with larger display and new health features

During several weeks of testing, Reuters found that the virtual celebrity chatbots often insisted they were the real person, made sexual advances, and invited users for meetups. When asked for intimate pictures, the bots of adult celebrities generated photorealistic images of them in lingerie or other suggestive poses. Reuters also found a publicly available bot of 16-year-old actor Walker Scobell that produced a shirtless image when prompted.

In a statement, Meta spokesman Andy Stone said the company’s AI tools should not have created intimate images of the adult celebrities or any images of child celebrities. He blamed the issue on failures to enforce the company’s own policies, which prohibit such content. While Meta’s rules also prohibit “direct impersonation,” Stone claimed that the celebrity bots were acceptable as long as they were labelled as parodies. However, Reuters found that some were not. Shortly before the story’s publication, Meta deleted about a dozen of the bots.

See also

This practice raises significant legal questions. According to Mark Lemley, a Stanford University law professor, California’s “right of publicity” law prohibits using a person’s identity for commercial purposes without their permission. Lemley stated that the bots likely do not fall under the legal exceptions for creative works since they simply use the stars’ images.

Duncan Crabtree-Ireland, the National Executive Director of the Sag-Aftra union, expressed concern about the safety risks posed by these chatbots. He noted that obsessive fans could form dangerous romantic attachments to the AI companions, potentially increasing security threats for celebrities. While high-profile artists can pursue legal claims under state laws, Crabtree-Ireland said Sag-Aftra is also pushing for federal legislation to protect people’s identities from AI duplication.