Government Recovers Record £480 Million in fraud, Deploying New AI to prevent Future Losses
LONDON – The UK government has recovered a record £480 million lost to fraud over the past 12 months (April 2024 – March 2025), according to the Cabinet Office. This marks the largest sum ever reclaimed by government anti-fraud teams in a single year, with £186 million specifically linked to fraudulent claims made during the COVID-19 pandemic. The recovered funds will be reinvested into public services, including the recruitment of nurses, teachers, and police officers, ministers announced.
The recovery efforts include preventing hundreds of thousands of companies with potentially fraudulent Bounce Back Loans from dissolving, thereby avoiding the loss of repayment. The Bounce Back Loan scheme, designed to support businesses during the pandemic with loans of up to £50,000, faced criticism for its lack of stringent oversight and vulnerability to fraud. One case identified by the Cabinet Office involved an individual who created a fictitious company and transferred loan funds to Poland.
To bolster future fraud prevention, the government has developed and is now deploying a new Artificial Intelligence (AI) tool, the “Fraud Risk Assessment Accelerator.” Developed by researchers within the Cabinet Office, the tool scans new policies and procedures for potential weaknesses before implementation, aiming to make them “fraud-proof.” Cabinet office minister Josh simons is announcing the savings and the tool’s rollout at an international anti-fraud summit with the US, Canada, and Australia on Wednesday.
“Cutting-edge AI and data tools” will be crucial in protecting public funds and preventing criminals from profiting, Simons stated. The UK government will license the tool for international use, with anticipated adoption by the US, Canada, australia, and New Zealand.
However, the move to expand AI-driven fraud detection raises concerns. Last year, an AI tool used by the Department for Work and Pensions to identify welfare fraud was found to exhibit bias based on factors like age, disability, marital status, and nationality, showing a “statistically meaningful outcome disparity” in a fairness analysis. Amnesty International recently criticized the government’s “unchecked use of tech and AI systems,” highlighting potential risks associated with such technologies.
