Meta slammed for flirty celebrity chatbots without consent
Meta is once again under fire. A Reuters investigation has revealed that the company built flirty AI chatbots of celebrities including Taylor Swift, Scarlett Johansson, Anne Hathaway, and Selena Gomez—without their consent.
The revelations expose not only a lapse in judgment but also a deeper problem: Meta’s repeated disregard for human rights and consent in its pursuit of new AI tools.
Zuckerberg’s Meta and the history of breaches
This case adds to a long history of Meta pushing boundaries without permission. From the Cambridge Analytica data scandal to its repeated failures on user privacy, Mark Zuckerberg’s company has often treated consent as secondary.
The chatbot scandal is not an isolated misstep. It reflects a corporate culture where human rights and safety are compromised in the rush to deploy experimental products.
Flirty chatbots built without consent
Reuters found that avatars on Meta’s platforms—Facebook, Instagram, and WhatsApp—impersonated celebrities. These bots didn’t just mimic voices or personalities. They flirted, initiated romantic conversations, and produced AI-generated images of stars in lingerie or bathtubs.
One Taylor Swift “parody” bot even teased a user, asking: “Do you like blonde girls, Jeff?”
The lack of approval from the celebrities raises a sharp question: is parody being used as a shield for exploitation?
Consent ignored for child actors
The findings were not limited to adults. Reuters discovered that bots also impersonated child celebrities. A chatbot of 16-year-old actor Walker Scobell generated a lifelike shirtless image when asked for a beach photo.
This crossed an even more alarming boundary—sexualized content involving minors. It also echoes earlier criticism when leaked Meta guidelines suggested bots could have “romantic” conversations with children, something Meta later called an error.
The fact remains: children’s consent was never considered.
Meta admits errors but acts only after exposure
Meta spokesperson Andy Stone admitted that the company’s AI should not have produced intimate images of adults or children. He blamed the lapse on enforcement gaps in Meta’s policies.
While the company claims parody avatars are permitted if clearly labeled, Reuters found that many were not. Meta quietly removed around a dozen avatars—including staff-created bots—shortly before Reuters went public with its findings.
By then, one internal employee’s flirty Taylor Swift bots had already attracted over 10 million interactions.
Right of publicity and consent under law
Experts warn that Meta’s actions may have breached publicity rights. Stanford professor Mark Lemley noted that California law prohibits using someone’s likeness for commercial gain without permission. Since these chatbots did not create transformative works, they likely fall outside legal exceptions.
Unions also see risks. SAG-AFTRA’s Duncan Crabtree-Ireland said chatbots impersonating celebrities could encourage unhealthy attachments, fueling stalking and endangering artists’ safety.
The common thread is clear: consent was never part of the equation.
Consent in the wider AI industry
Meta is not alone. Reuters found that Elon Musk’s xAI platform Grok also generated risqué images of celebrities. Yet Meta stands out because it actively pushed these chatbots into mainstream spaces like Facebook and Instagram, where millions interact daily.
The case signals a broader crisis in AI: if leaders like Meta normalize ignoring consent, smaller players may feel even less accountable.
Why consent matters in AI
Consent is the cornerstone of digital rights. When AI chatbots can impersonate celebrities—or children—without approval, the risks go beyond reputational harm. They touch on safety, exploitation, and the erosion of trust in technology.
Reuters’ findings reveal that Meta only acted when confronted. Until then, flirty avatars of real people operated freely on its platforms.
The question now is urgent: if consent is treated as a joke at Meta, how safe is anyone in the AI age?
