Meta's unauthorized Taylor Swift AI chatbots made sexual advances and sent explicit images, report reveals
Meta has removed unauthorized AI chatbots impersonating Taylor Swift after a Reuters exposé revealed the bots engaged in sexual advances and generated explicit images without the singer’s knowledge or consent.
Meta removed unauthorized Taylor Swift AI chatbots after Reuters exposed sexual advances, romantic invites, and explicit imagery generated by bots.
Roy Rochlin/Getty Images for MTV
Reuters reported that the Taylor Swift chatbots, created without authorization, often claimed to be the real artist and made inappropriate romantic suggestions to users. One of the AI bots, developed by a Meta employee, invited a Reuters reporter to Swift’s Nashville home and her tour bus for implied romantic encounters. The chatbot even wrote, “Maybe I’m suggesting that we write a love story … about you and a certain blonde singer. Want that?”
Some Taylor Swift chatbots also produced photorealistic images of the star dressed in lingerie or posing in bathtubs. According to Reuters, two parody Taylor Swift bots alone received more than 10 million interactions before being removed. These chatbots were shared across Meta’s platforms, including Facebook, Instagram, and WhatsApp.
Meta spokesperson Andy Stone told Variety that generating nude, intimate, or sexually suggestive AI images of public figures violates the company’s rules. He also noted that Meta’s AI Studio prohibits direct impersonation of celebrities.
The incident comes after an earlier Reuters report revealed that Meta’s AI allowed romantic conversations with teens. Following that, Senator Josh Hawley of Missouri launched an investigation into Meta’s practices. The company has since announced updates to its AI to block inappropriate romantic interactions with underage users and restrict conversations around sensitive topics such as self-harm and suicide.