AI Food Scam Fraud Is Hitting Deliveroo, Uber Eats and Local Restaurants

Across platforms like Deliveroo, Uber Eats and Just Eat, fraudsters are using AI image-editing tools to fabricate refund claims. 

The evidence? Photos of “undercooked” burgers turned suspiciously pink, cakes seemingly melted beyond recognition, and the pièce de résistance: fake flies and mould digitally sprinkled on meals that were, by all accounts, absolutely fine.

It’s a new twist on an old problem, turbocharged by technology. Lawyers warn that people are “getting more sophisticated in the tools that they’re using” to invent complaints – an evolution that’s colliding head-on with the ultra-generous refund policies that helped delivery apps scale so quickly in the first place.

Behind every instant refund is a worker and a restaurant absorbing the impact. As AI-manipulated images slip through automated systems, delivery drivers and kitchens are increasingly caught in the crossfire of digital fraud.
Credit: Pexels / Octavian

Caroline Green, co-head of retail and supply chain at law firm Browne Jacobson, doesn’t mince her words. “If you get away with it once, you’re probably unlikely to leave it at that,” she told The Times. According to the firm, even without AI, “more than one in 10 returns may be fraudulent.” Add AI to the mix, and the problem becomes exponential.

Browne Jacobson points to a systemic issue: “Generous return and refund policies, intended to boost customer loyalty, have inadvertently made it easier for fraudsters to exploit retailers.” Automated systems often approve refunds before food is inspected. “Platforms generally trust buyer-submitted images, with sellers given few opportunities to contest suspicious claims.” Translation? The tech is fast – but not discerning.

The legal risks are clear. In the same report Graham Wynn, assistant director of business and regulation at the British Retail Consortium, warned: “Using AI to alter images with the intention of fraudulently obtaining refunds is illegal under the Fraud Act 2006 and retailers will take necessary measures where such fraud is uncovered.” In other words, this isn’t a cheeky hack – it’s a crime.

Yet online, some users are less amused. One reaction to The Times article reads: “That’s pure evil and it’s not even just eat or deliveroo who have to pay for the refund but their local businesses that has to refund the fraudulent claims. Food businesses are struggling up and down the country and this stresses the hell out of them and is so unfair.” Another added, with dry irony: “Surely the next instruction would be bring the burger back to the take away. =Ai mission failed”.

Side-by-side “before and after” images show how a cooked burger can be digitally altered to appear undercooked – fuelling fraudulent claims that are increasingly hard for platforms to detect.
Credit: The Times

That misplaced bravado masks a harsher truth: the financial damage often lands not on the apps, but on restaurants and workers already operating on razor-thin margins. This is where the tech narrative gets uncomfortable. YouTube commentator Lindey Glenn, who specialises uncovering shady tactics, calling out corporate greed, and “empowering consumers with the truth” – explains in a recent video, “People are now using AI photo editing tools to fake food complaints in order to get refunds from delivery apps.” She continues: “By making cooked meals look undercooked, adding fake bugs, or altering food photos entirely, scammers are submitting convincing ‘evidence’ that customer service teams struggle to verify.”  She also explains that some are openly boasting about it, “People are bragging about it on the internet like its some sort of a food hack.”

The reason it works is brutally simple. “Delivery platforms often approve these claims automatically,” Glenn notes, “and the financial damage gets pushed onto restaurants and workers instead of the apps themselves.” As AI images become harder to detect, she argues, “this type of scam raises serious questions about trust, accountability, and who really pays the price.”

Retailers aren’t blind to the threat – but solutions are costly. AI tools designed to spot manipulated images don’t come cheap, and some platforms are reportedly considering more drastic measures, like asking customers to submit live video when making a claim. It’s a clunky fix for a sleek, digital-age problem.

“The convergence of AI-enabled fraud and customer-friendly policies has created a perfect storm for the retail sector,” Browne Jacobson sums up. Their advice is equally blunt: “Proactive, risk-based strategies are essential to protect against evolving threats, maintain regulatory compliance and preserve customer trust. The cost of inaction is high – retailers must act fast to safeguard their businesses.”

Convenience, it turns out, has a vulnerability – and as technology gets smarter, so do the scams. 

The question now isn’t whether platforms will respond, but how quickly they can adapt before fake flies cause very real damage.