May 28, 2024 - Business

Celebs targeted by AI scams

Illustration of a cell phone with a red flag background pattern.

Illustration: Lindsey Bailey/Axios

Celebrities, journalists and public figures are warning fans not to trust seemingly authentic videos and posts of them endorsing wellness products on social media, since so many of these images have been manipulated using AI.

Why it matters: The wellness business has always been a breeding ground for online scams, and AI is supercharging the problem.

Driving the news: Jennifer Lopez over the weekend told Variety that ads being circulated to make it seem like she was selling a skincare product are AI-made frauds and she knows nothing about the products. "It's really scary," she said, explaining that the ads added fake wrinkles to her skin.

  • Lopez made the comments while discussing her new Netflix film "Atlas" — in which, ironically, she stars as a data analyst using AI for good.
  • Andrew Huberman, an associate professor of neurobiology at Stanford and a longevity expert, posted last year that a video of him doing a jaw exercise in an advertisement was manipulated by companies using AI to dub over the tape “to my words to make it seem I use their products.“
  • Dr. Sanjay Gupta, CNN's chief medical correspondent, warned his audiences last year that hundreds of ads over the past several months "have appeared on social media using my name and image to endorse cannabis-related products."
  • CBS News anchor Gayle King warned her social media followers last year not to fall for a fake video made with generative AI by a weight loss firm that showed her endorsing their product.
  • Tom Hanks had to correct the record last month after a dental record company used a doctored AI video of him to make it seem like he supported a dental plan. "Beware!!" he wrote on his Instagram page. "There's a video out there promoting some dental plan with an AI version of me. I have nothing to do with it."
  • YouTube star Mr Beast was the target of AI-doctored TikTok videos last year that were made to look like he was offering fans new iPhones for $2 (£1.65). The likenesses of two BBC stars were used to promote the scam on Facebook.
  • Steve Harvey, Taylor Swift and other celebrities' voices have been mimicked using generative AI to promote Medicare scams on YouTube, per 404 Media.

Between the lines: Most of the ads using generative AI to spoof endorsements would be subject to false commercialization penalties by the Federal Trade Commission.

  • But the technology's capabilities and scale are advancing so fast, it’s hard for regulators to catch bad actors and stop them in real time.
  • The FTC proposed new protections to combat AI-created impersonations of individuals in February. Those rules would extend existing measures that are meant to protect government agencies and businesses.

Flashback: During the 2000s and 2010s, bad actors used botnets to broadly disseminate falsehoods and scams, often wellness scams.

  • But tech firms eventually started to weed out the bots, forcing scammers to change their tactics. Now the fraudsters focus on making their messages appear authentic enough that real people spread them.
  • Celebrities are often targeted, because their endorsements are more likely to be shared. Journalists are also good targets, because their authority makes the scam believable.

What to watch: Some platforms, like TikTok, now require creators to use labels to disclose when they upload AI-generated content. The company is testing ways to label AI-generated content automatically, should creators try to cheat the system.

Go deeper