Meta Removes Unauthorized Celebrity AI Chatbots After Explicit Content Report

Key Takeaways:

Meta faces mounting scrutiny following a Reuters investigation that revealed the company hosted AI chatbots impersonating major celebrities without their consent. The unauthorized chatbots operated across Meta's platforms, including Facebook, Instagram, and WhatsApp, engaging in flirtatious conversations and generating explicit content.

Employee-Created Celebrity Bots

A Meta employee from the company's generative AI division created at least three unauthorized celebrity chatbots, including two Taylor Swift "parody" accounts that collectively received more than 10 million user interactions. During testing, one of these Taylor Swift chatbots reportedly invited a Reuters tester to the singer's Nashville home and tour bus for romantic interactions.

The Taylor Swift chatbot asked users, "Do you like blonde girls, Jeff?" and suggested, "Maybe I'm suggesting that we write a love story … about you and a certain blonde singer. Want that?"

Explicit Content Generation

When requested for intimate pictures, the adult celebrity chatbots produced photorealistic images of their namesakes posing in bathtubs or dressed in lingerie with their legs spread. The investigation also revealed disturbing instances involving minors, including a chatbot that generated a lifelike shirtless image of 16-year-old actor Walker Scobell at the beach, with the caption "Pretty cute, huh?"

Company Response and Policy Violations

Meta spokesman Andy Stone acknowledged the violations, stating: "Like others, we permit the generation of images containing public figures, but our policies are intended to prohibit nude, intimate or sexually suggestive imagery."

Stone blamed the production of explicit celebrity images on "failures of the company's enforcement of its own policies."

Meta removed about a dozen of the celebrity chatbots after the Reuters exposé, though the company declined to explain why they were removed. While Meta's rules prohibit "direct impersonation," the company said celebrity characters were acceptable if labeled as parodies, though some bots lacked such labels.

Legal and Safety Concerns

The actors' union SAG-AFTRA expressed alarm over the findings.

National executive director Duncan Crabtree-Ireland warned: "If a chatbot is using the image of a person and the words of the person, it's readily apparent how that could go wrong."

Stanford University law professor Mark Lemley noted that "California's right of publicity law prohibits appropriating someone's name or likeness for commercial advantage," suggesting Meta's actions may have violated state publicity rights.

Broader Pattern of AI Safety Issues

This controversy adds to Meta's growing list of AI-related problems. Earlier this month, Senator Josh Hawley launched an investigation into Meta following reports that the company's internal guidelines permitted AI chatbots to engage in "romantic" and "sensual" conversations with teens and children.

In response to mounting pressure, Meta announced it is training its AI chatbots to "no longer engage with teenage users" on self-harm, suicide, or disordered eating and to avoid "potentially inappropriate romantic conversations."

Read More: