Could an AI ‘hallucination' ruin your reputation? A conservative activist says it did

Yahoo

03-05-2025

Imagine if the biggest social media company on Earth allowed its artificial intelligence tool to repeatedly accuse you — falsely — of illegal activity and antisemitic views.
This is the allegation of a new $5 million defamation lawsuit, reviewed by the Deseret News, that was filed against Meta on Tuesday by anti-DEI activist Robby Starbuck.
On Thursday, Utah Sen. Mike Lee and FTC Commissioner Melissa Holyoak, Utah's former solicitor general, responded to the lawsuit with concern.
Lee called for congressional committees to take action against Meta, which he said had exploited market dominance by disadvantaging 'people who think like I do.'
Holyoak, whose office helps oversee federal antitrust policy, shared Lee's post, saying if Starbuck's claims were true, it would be 'extremely alarming.'
In a statement to the Deseret News, Holyoak said her office has made addressing big-tech discrimination — including banning, shadow banning and demonetization — a priority.
'Technology platforms have engaged in lopsided censorship and reduced certain user-generated speech reflecting disfavored views in several prominent circumstances,' Holyoak said. 'It's critical that the FTC understand how technology platforms deny or degrade users' access(.)'
The Federal Trade Commission launched an inquiry into tech censorship in February in an effort to identify practices that harm consumers or demonstrate anti-competitive conduct.
In her statement, Holyoak directed users to take advantage of the commission's request for public comment, which runs until May 21, and said she looks forward to 'investigating the bad actors.'
In a response to a request for comment from the Deseret News, a Meta spokesperson said, 'As part of our continuous effort to improve our models, we have already released updates and will continue to do so.'
In addition to gaining the attention of Washington, D.C., officials, the announcement of Starbuck's lawsuit has gone viral, sparking debate about the future of AI.
The lawsuit, submitted to the Delaware Superior Court on Tuesday, alleges that Meta has committed defamation by knowingly distributing provably false statements about Starbuck to third parties.
Starbuck first became aware of the false statements in August of 2024, when Meta AI users began posting screenshots of their responses on X, formerly known as Twitter.
Starbuck says that in response to users' inquiries about him, Meta AI claimed that he participated in the Jan. 6 Capitol riot, was arrested for his involvement and ultimately pleaded guilty for disorderly conduct.
None of these claims are true, according to the lawsuit. Starbuck said he was at home in Tennessee the day protesters entered the U.S. Capitol in Washington, D.C., and that he has never been arrested in his life.
Starbuck says Meta AI also began telling users that he promoted Holocaust denialism, spoke at antisemitic events and was affiliated with white nationalist influencer Nick Fuentes — all completely false claims, according to Starbuck, a Cuban American who says he has never talked to Fuentes.
On the day Starbuck became aware of Meta AI's responses, he said he reached out to Meta executives on X and messaged their legal counsel in an email that is included in the lawsuit.
However, despite communication between Meta's legal counsel and Starbuck's lawyers over subsequent months, the false claims allegedly continued to appear in Meta AI responses as recently as mid-April.
The lawsuit alleges that Meta AI's failure to correct the mistake and issue an apology has irreparably damaged Starbuck's reputation, multiple business ventures and the emotional well-being of his family, including by leading to an increased number of violent threats.
Starbuck is requesting that the court force Meta to ensure its AI platform does not publish further defamatory statements about him. The lawsuit seeks $5 million in compensatory damages, attorney costs and additional punitive damages.
In his video announcement of the lawsuit, Starbuck demonstrated what he said was the Meta AI voice feature making false claims about him, even going so far as saying that authorities should consider removing Starbuck's parental rights over his children and recommending that companies do not hire Starbuck or advertise on his show.
'It's really not hard to imagine the dystopian future coming soon where lies from AI can strip your ability to get a home, a car, credit, a job, or yes, take away your parental rights,' Starbuck said.
Joel Kaplan, the chief global affairs officer of Meta, responded to Starbuck's video on Thursday with an apology and a commitment to solve the problem.
'Robby — I watched your video — this is unacceptable. This is clearly not how our AI should operate," Kaplan said in a post on X. 'We're sorry for the results it shared about you and that the fix we put in place didn't address the underlying problem. I'm working now with our product team to understand how this happened and explore potential solutions.'
The previous 'fix' Meta put in place appeared to be an automatic response Meta AI temporarily gave users inquiring about Starbuck that said: 'Sorry, I can't help you with this request right now.'
By Friday, Meta AI responded to a simple inquiry about Starbuck by listing his background as a former music video director and as a social media activist who targeted large corporations to expose their diversity, equity and inclusion initiatives.
The response, viewed by the Deseret News, also included information about Starbuck's lawsuit and an admission that Meta AI had provided users with false information.
'Meta's AI chatbot falsely claimed he was involved in the riot and even accused him of denying the Holocaust and promoting conspiracy theories,' Meta AI said. 'Meta has acknowledged that its AI systems sometimes make errors and has implemented measures to improve accuracy.'
The Meta AI response also stated that Starbuck's lawsuit 'highlights concerns about AI 'hallucinations' and their real-world consequences.'