AI Brad Pitt dupes woman out of €830K

If a Hollywood superstar slid into your DMs asking for money, you would probably block them instantly. But what if they backed it up with exclusive, unsearchable hospital selfies and a custom AI-generated TV news segment confirming your secret relationship?

That was the trap sprung on Anne, a 53-year-old French interior designer who lost €830,000 ($850,000) to scammers impersonating Brad Pitt over an 18-month ordeal. The story went viral after French broadcaster TF1 aired a primetime segment on Sunday, but instead of sympathy, Anne was met with overwhelming online mockery, prompting the network to pull the clip.

Anatomy of a high-tech romance scam

Anne's nightmare began in February 2023, shortly after she downloaded Instagram for the first time while navigating a failing marriage.

The scammers initially reached out posing as Pitt's mother, Jane Etta, claiming her son needed a woman just like Anne. The next day, a profile claiming to be Pitt took over.

Every time Anne doubted the situation, the scammers deployed tailored excuses and generative tech. First came a request for €9,000 to cover customs on luxury gifts, allegedly held up because Pitt's bank accounts were frozen due to his divorce from Angelina Jolie.

Then things escalated. The fake Pitt claimed he was battling kidney cancer and needed financial help for treatment, sending multiple AI-generated photos of himself in a hospital bed. Because reverse image searches yielded zero matches online, Anne assumed the photos were real private selfies taken just for her. Anne, herself in cancer remission, felt she was saving a life.

She eventually divorced her wealthy husband and wired her entire €775,000 settlement to the scammers, despite her 22-year-old daughter pleading with her for over a year to see reason.

When celebrity magazines reported on the real Pitt dating Ines de Ramon, the fraudsters countered with a fake TV broadcast featuring an AI news anchor detailing Pitt's exclusive relationship with a special woman named Anne. The illusion finally shattered in June 2024 when Pitt officially went public with his girlfriend. By the time scammers tried extorting more cash posing as FBI Special Agent John Smith, Anne went to the police.

The internet chooses mockery over empathy

Anne was left penniless, living out of cardboard boxes in a friend's room, and reported attempting suicide three times since the scam unraveled.

Yet online users turned her misery into a punchline. Brands jumped in as well: Netflix France posted on X promoting four films with Brad Pitt (for real), while soccer club Toulouse FC posted a mocking tweet before taking it down and apologizing.

A representative for Pitt issued a statement calling it awful that scammers exploit fans' connections with celebrities, reiterating that people should ignore unsolicited outreach, especially from stars who do not even have social media accounts.

Speaking to a popular French YouTube show, Anne defended herself, stating she was neither crazy nor an idiot, just someone who got played and wanted to speak out because she is far from the only victim.

Why it matters

This case demonstrates how generative AI has fundamentally broken traditional scam-detection reflexes.

For years, standard safety advice for online dating was simple: run a reverse image search. But generative models render that advice obsolete. Because the hospital photos were fabricated from scratch, the lack of search results perversely served as proof of authenticity to Anne. Combined with AI-generated news broadcasts, scammers now possess the tools to construct personalized, highly persuasive parallel realities.

We think the incident also highlights a troubling gap in digital literacy and public empathy. Younger netizens often laugh off these scams as obvious absurdities, forgetting that deepfakes are uniquely weaponized against older or emotionally vulnerable individuals who did not grow up navigating online deception.

Mocking victims only discourages others from coming forward, allowing these criminal networks to operate undisturbed while generative AI makes their synthetic traps increasingly impossible to spot.