The UK’s tax authority is turning to AI to help identify fraud

HM Revenue & Customs (HMRC) has announced a 10-year, £175 million ($234 million) partnership with British tech firm Quantexa to deploy artificial intelligence (AI) for detecting tax fraud and errors. This initiative aims to strengthen the UK tax authority’s ability to identify irregularities in tax filings and payments, marking a significant step in government adoption of AI technology for public finance management.

The agreement, announced in March 2024, involves Quantexa integrating its AI data analytics tools with HMRC’s existing data sources. The technology will assist in identifying both deliberate fraud and unintentional errors in tax returns, as well as helping to track down taxpayers who use incorrect reference numbers for legitimate payments. Quantexa, founded in London in 2016, develops AI applications focused on data analysis and decision-making. Its systems will support HMRC staff by flagging potential issues for further human review, rather than making autonomous decisions.

According to Quantexa’s CEO Vishal Marria, the company’s AI is designed to support transparency, auditability, and explainability, especially in government contexts where decisions directly affect citizens. Marria emphasized that the data will remain within HMRC’s environment, ensuring security and compliance with privacy standards. The company also highlighted that its AI tools are used in other sectors, such as Zurich Insurance Group’s fraud detection efforts in Switzerland.

Why It Matters

This development signifies a broader shift among governments toward integrating AI into tax enforcement and fraud prevention. By leveraging advanced data analytics, HMRC aims to improve detection accuracy, reduce fraud, and enhance taxpayer compliance. The move could lead to more efficient resource allocation and potentially reduce financial losses due to tax evasion. For taxpayers, this may mean increased scrutiny but also potentially fewer false positives, as AI tools are intended to support human judgment.

Furthermore, the deal reflects growing confidence in AI’s role within public sector operations, following similar moves by the US Treasury Department, which reported recovering over $4 billion using AI in the 2023-2024 period. The UK’s approach underscores a trend toward transparency and security in AI deployment for sensitive government functions.

Background

In recent years, governments worldwide have increasingly adopted AI technologies for financial and security purposes. The US Treasury Department, for example, announced in 2024 that it used AI to recover billions in unpaid taxes and prevent fraud. The UK’s HMRC has historically relied on traditional data analysis methods, but this partnership signals a strategic shift toward more advanced, automated systems to keep pace with sophisticated tax evasion tactics.

Quantexa, which has also partnered with Zurich Insurance to improve fraud detection, is part of a growing industry of AI firms providing government and private sector solutions. The UK government’s investment in this technology aligns with broader efforts to modernize public services and improve compliance through digital transformation.

“”In government environments, AI cannot operate as a black box. Decisions need to be transparent, auditable, and explainable, particularly in areas affecting citizens directly.””

— Vishal Marria, Quantexa CEO

“”We will never take HMRC data away from the HMRC environment to ensure the data remains secure.””

— Vishal Marria, Quantexa CEO

What Remains Unclear

It remains unclear how the AI system will perform in practice, including its accuracy and the rate of false positives. The exact scope of AI influence on individual cases and the level of human oversight are still to be clarified. Additionally, the long-term impact on taxpayer behavior and compliance is yet to be observed, as the initiative is in its early stages.

What’s Next

In the coming months, HMRC is expected to begin phased implementation of the AI tools, with ongoing monitoring of their effectiveness. Further updates on performance metrics, adjustments based on initial results, and potential expansion of AI capabilities are anticipated. The government may also publish evaluations to address transparency and public trust concerns.

H&R Block Tax Software Deluxe + State 2022 with Refund Bonus Offer (Amazon Exclusive) [PC Download] (Old Version)

Choose to put your refund on an Amazon gift card and you can get a 2.75% bonus. See…

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

How will AI improve tax fraud detection?

AI will analyze large datasets to identify patterns and anomalies indicative of fraud or errors, enabling HMRC to target investigations more effectively.

Will AI lead to false accusations against taxpayers?

The AI is designed to support human decision-making, not replace it. HMRC officials will review flagged cases to prevent wrongful accusations.

Is taxpayer data safe in this AI system?

Yes, Quantexa has stated that all data will remain within HMRC’s secure environment, ensuring compliance with privacy standards.

How long will the AI system be used?

The partnership is set for 10 years, with ongoing assessment and potential updates during this period.

Are other countries using AI for tax enforcement?

Yes, the US Treasury Department has reported using AI to recover billions in unpaid taxes, and other nations are exploring similar technologies.