Is the UK government’s new AI tool preventing fraud?
MKS Law Head of Investigations doubts whether AI alone can prevent massive fraud taking place in the UK.
The UK government has credited a high-tech surge, driven by artificial intelligence (AI) and advanced data analytics, with delivering the country’s “biggest ever fraud crackdown,” saving the taxpayer a record £480m ($640m) in the year to April 2025.
The Fraud Risk Assessment Accelerator (FRAA) has been heralded as the future of the UK government’s anti-fraud strategy. Unlike previous detection tools that identify fraud after it occurs, the FRAA is designed to stop large-scale theft before new policies are even rolled out, said the Cabinet Office this week.
The FRAA is a generative AI system built to scan draft government policies and procedures, pinpointing weaknesses that sophisticated fraudsters could exploit.
Results from early testing show the tool could reduce the time needed to identify fraud risks by up to 80%, potentially saving millions in future losses by making initiatives fraud-proof from the outset. This preventative capability is a direct response to the massive losses seen during the COVID-19 pandemic, where emergency loan schemes were exploited by criminals.
Yet our Head of Investigations, Tony McClements, has doubts, as he told GRIP.
“Although AI is making a very small dent in attempted fraud, human resources are an issue. Without trained officers – that is, boots on the ground – AI will have limited impact. We should not be satisfied and patting ourselves on the back for preventing some of the fraud: we need to go after the culprits and prosecute them. Prosecutions create consequences, which in turn creates deterrent. Without deterrents, the criminals simply carry on unabated.”
McClements further criticised the established structural response to large-scale criminality, specifically taking aim at the National Anti-Fraud Taskforce (NATIS) in the wake of the pandemic fraud issue.
“What a shambles, you couldn’t make this up. I wrote in the aftermath of the COVID fraud debacle that a specialist squad, made up of serving and retired detectives, was the way forward. NATIS has failed miserably, and the fraudsters are laughing all the way to the bank. We cannot walk away from a crime of this magnitude simply because it is too-hard-to-do and those charged with investigating and prosecuting it have been proven thus far incompetent.”
Both the government’s record crackdown and the financial sector’s pioneering AI pilots underline a fundamental shift in the UK’s approach to economic crime: moving from reactive loss recovery to proactive, AI-driven prevention across both public and private spheres.
However, as Tony McClements laments, without the deterrent of prosecutions, “the criminals simply carry on unabated”.
