New research reveals deepfake fraud cases surged by almost 1,000 per cent in 2025, with celebrities increasingly used to promote fake investments, cryptocurrency schemes and online scams…
The growing threat posed by artificial intelligence-powered scams has been laid bare by new research showing deepfake fraud incidents skyrocketed by an astonishing 982 per cent in 2025.
According to analysis by online broker BrokerChooser, criminals are increasingly exploiting the images and voices of celebrities, politicians and business leaders to deceive victims, with AI-generated content now contributing to an estimated £1.8 billion in annual losses to the UK economy.
The findings come amid mounting concern over the sophistication of deepfake technology and the public’s ability to distinguish fake content from genuine videos. Recent studies suggest around 70 per cent of UK adults struggle to reliably identify AI-generated videos, while one in five people correctly identified only a single deepfake clip when tested.
Donald Trump most targeted by deepfake scammers
The research, which examined more than 4,000 historical AI-related fraud incidents, found that Donald Trump was by far the most frequently impersonated public figure, appearing in almost one in five reported deepfake cases.
Trump accounted for 19.58 per cent of all incidents analysed, making him the clear favourite among fraudsters seeking to exploit public recognition and trust.
In second place was Elon Musk, who featured in 10.49 per cent of reported cases. Musk has become a particularly popular figure for investment and cryptocurrency scams, with convincing fake videos often portraying him endorsing fraudulent trading platforms and digital assets.
Actor Brad Pitt ranked third, accounting for 3.26 per cent of incidents, while pop superstar Taylor Swift was the most targeted musician, appearing in 3.03 per cent of cases.
Other prominent figures regularly impersonated included Scarlett Johansson, Tom Cruise, Joe Biden, Kamala Harris, Kanye West, Barack Obama and Mark Zuckerberg.
Celebrity trust exploited for financial fraud
Experts say the rise in celebrity deepfakes reflects a simple reality: people are more likely to trust a familiar face.
The Advertising Standards Authority has previously identified fake advertisements featuring celebrities and public figures as the most common type of scam advert appearing online.
Many of these scams follow a familiar formula. A convincing AI-generated video appears to show a trusted celebrity endorsing an investment platform, cryptocurrency scheme or financial product. Victims are then directed towards fraudulent websites designed to steal money or personal information.
The study found that more than half of all deepfake incidents recorded in 2025 involved celebrities, highlighting how central famous faces have become to modern online fraud.
Romance scams and emotional manipulation
While investment scams remain the dominant threat, some deepfakes are being used in more personal and emotionally devastating ways.
Brad Pitt’s image has become a frequent tool in romance scams, with fraudsters using AI-generated content to convince victims they are communicating with the Hollywood actor.
One widely reported case saw a woman allegedly lose around £700,000 after being convinced she was in an online relationship with Pitt, leaving her financially ruined and contributing to the breakdown of her marriage.
Meanwhile, Taylor Swift’s enormous global popularity has made her a prime target for scammers seeking viral reach. Fake videos featuring the singer have been used to promote fraudulent services on social media platforms, including bogus payment schemes and online giveaways.
Deepfake growth shows no sign of slowing
The research highlights an alarming acceleration in deepfake activity over recent years.
After relatively low levels of activity in 2021 and 2022, incidents surged by 750 per cent in 2023 as generative AI tools entered mainstream use. That was followed by a further rise of nearly 120 per cent in 2024 before the extraordinary 982 per cent jump recorded in 2025.
The figures suggest that as AI tools become cheaper, faster and more accessible, criminals are rapidly adopting the technology to scale up fraud operations.
How to spot a deepfake
Ádám Nasli, Head Broker Analyst at BrokerChooser, warned that while scams are becoming increasingly sophisticated, there are still ways consumers can protect themselves.
“People are naturally more likely to trust a familiar face, which is what makes ‘celeb-baiting’ such an effective tactic,” he said.
He advises consumers to be especially cautious of celebrity endorsements linked to investment opportunities, noting that legitimate firms operating in the UK must be authorised by the Financial Conduct Authority.
Experts also recommend paying close attention to subtle visual clues that may reveal AI-generated content. Common warning signs include unnatural blinking, awkward facial expressions, unusual jaw movements and lips that appear slightly out of sync with speech.
Watching videos more than once or slowing playback can often make these inconsistencies easier to spot.
