HMRC to use AI from British tech firm to spot fraud and tax return errors

Britain's tax authority HMRC has awarded a £175 million contract to Quantexa, a London-based financial data platform, to deploy artificial intelligence systems designed to detect fraud and spot errors in tax returns at scale. The multi-year deal represents a significant expansion of AI adoption within government operations and marks a major win for the British fintech sector.

Quantexa specializes in entity resolution and financial crime detection using machine learning. The company's platform analyzes patterns across massive datasets to identify suspicious transactions, compliance gaps, and filing mistakes that human reviewers might miss. For HMRC, this translates into faster identification of tax evasion schemes, undisclosed income, and computational errors that cost the British exchequer billions annually.

The contract underscores the UK government's pivot toward AI-driven compliance infrastructure. Rather than hiring additional auditors or tax inspectors, HMRC will leverage algorithmic screening to flag high-risk cases for human investigation. This hybrid approach allows the tax authority to process more returns with existing staff while theoretically improving accuracy.

The deal also signals confidence in homegrown British tech talent. Quantexa, which has raised substantial venture funding and counts major financial institutions among its clients, competes globally against established players like Thomson Reuters and SAS Institute. Winning a government contract of this scale validates the company's technology and positions it as a serious contender in the financial intelligence space.

Industry observers note this mirrors broader government trends toward outsourcing complex technical work to private firms rather than building capabilities in-house. Whether the AI system will actually recover more tax revenue remains an open question, as does public sentiment around automated fraud detection. The rollout will likely face scrutiny from privacy advocates concerned about algorithmic decision-making affecting taxpayers.