BBC Finds AI-Generated Fitness Ads Breaching UK Rules

Polished fitness videos promising dramatic body transformations in weeks have flooded social media. Chiselled physiques, striking before-and-after images and claims of looking years younger through simple routines often appear too good to be true. In many cases, they are.

A BBC investigation revealed misleading fitness adverts featuring AI-generated characters that breach UK advertising rules. Many failed to disclose that the featured people were not real. The ads aimed to sell subscriptions to a fitness app.

AI content has inundated social media feeds in recent years. Videos promoting exercise and online fitness programs are increasingly common. Adverts flagged to the Advertising Standards Authority by the BBC showed AI-generated characters claiming to have followed workout programs. They displayed transformations experts called scientifically implausible in short timeframes.

The videos promised body changes in weeks, to "look 20 years younger," or to "lose 40lb in one month." Algorithms flood users' feeds with similar material after they engage with fitness content.

Prof Andy Miah, an AI expert from the University of Salford, called the trend "huge." He said people seek solutions for health, fitness and looks, creating demand for such content. "There's always been an appetite for that kind of content - but now it's incredibly hard to tell who to believe."

Unlike human influencers, AI characters produce content endlessly, and users cannot opt out. "You can't turn [AI content] off," Prof Miah said. "It's impossible to stop your feeds being proliferated with this material."

He noted positive aspects to AI but called the current regulation a "wild west." Some ads could harm viewers. "The claims about how quickly you can make gains are completely unrealistic," he said. "That feeds false hope and creates damaging expectations."

The BBC contacted companies behind problematic adverts. None responded.

On a beach in North Tyneside, fitness instructor David Fairlamb led a group training session for nearly 40 people of all ages. He has worked in fitness for 30 years, before social media and AI.

Fairlamb, 54, said AI has a place in fitness and nutrition but cannot replace real coaching. "You cannot beat that real person, that real connection, the accountability."

Shown the breaching AI ads, he reacted strongly. "It's so wrong. It's so misleading. And it's so worrying for younger kids."

"These ads talk about 28-day transformations. I've been doing this for 30 years and I'm telling you now - that just doesn't happen. You've got no chance."

Fairlamb works with his daughter Georgia Sybenga, 25. She said even social media natives struggle to spot fakes. "Sometimes I question it myself," she said. "Some of them, you really can't tell."

Both worry constant exposure to idealized artificial bodies harms confidence, especially among youth. "They think 'I could look like that in 30 days'," Fairlamb said. "But that body might not even be real. For young lads, for their mental health, it's really concerning."

Sybenga warned AI programs ignore injuries or health conditions. "It doesn't take into consideration injuries or health conditions, so... you could injure yourself."

The ASA said AI is not banned in ads but messages must not mislead or harm. "We don't judge ads based on whether they contain AI. We judge them on whether they're misleading or likely to be harmful," said Adam Davison, the ASA's director of data science.

The regulator received about 300 AI ad complaints in the past year, with numbers rising. "One challenge is that sometimes it can be hard even for us to tell whether AI has been used in an ad," Davison added.

AI tools enable quick ad generation, often by those unfamiliar with rules. The ASA issued "advice notices" to BBC-flagged advertisers with unsubstantiated claims. No prior complaints existed against them, so the BBC withheld identities.

"A big part of what the ASA does, as well as our enforcement work, is trying to educate advertisers on their responsibilities," Davison said. "If you're not being careful to review the content that's coming out of those tools then it's very easy to have something misleading ending up being posted."

Social media firms say AI content should be labelled. The BBC found hidden, unclear or missing disclaimers. Meta and TikTok declined comment.

TikTok has labelled over 1.3 billion AI-generated videos. Meta uses indicators from creation tools.

Users the BBC spoke to wanted an opt-out for AI content. Meta and TikTok did not say if considering it.

Prof Miah said social media economics favor more AI content. "It's clearly useful in many ways. But where it then misleads people to have false expectations... is where perhaps regulation needs to step in."