UK government says a new AI tool helped it recover almost £500 million in fraud losses – and now it’s going global
- AI fraud detection system helped UK government recover record £480m in one year
- It targeted Bounce Back Loan fraud and prevented thousands of suspicious firms from dissolving
- Civil liberties campaigners warn against growing use of AI in government
The UK government says a new artificial intelligence system has helped its anti-fraud teams recover almost £500 million over the past year, the largest amount ever reclaimed in a 12-month period.
The Cabinet Office put the total reclaimed between April 2024 and April 2025 at £480m. Of that figure, around £186m related to fraudulent activity tied to the Covid pandemic, when support schemes such as the Bounce Back Loan programme were introduced at speed.
Worth up to £50,000 each, those loans were supposed to help companies struggling as a result of lockdown, but they were widely criticized for being given out without proper checks.
Stopping “scammers and swindlers”
The BBC reports ministers now plan to put the savings into frontline services such as health, education and policing.
The new AI system, called the Fraud Risk Assessment Accelerator, helped officials prevent thousands of firms with suspect loans from dissolving to avoid repayment.
The tool, which was developed by government researchers, can review proposed policies and spot weak points before they are taken advantage of.
Cabinet Office minister Josh Simons is expected to announce the recovery figures at an anti-fraud summit co-hosted by the UK, the US, Canada and Australia.
He said the government wanted to stop “scammers and swindlers” from draining public funds.
The BBC reports that following its successful outcome, the tool will be licensed internationally, with countries including the US, Canada, Australia and New Zealand expected to make use of it.
Not everyone will be thrilled about the Fraud Risk Assessment Accelerator’s involvement in recovering fraudulently claimed funds.
Plenty of people are worried about the increasing use of AI tools in government and claims of bias and unfair outcomes were raised by civil liberties groups last year after similar technology was used to check welfare claims.
