Quantexa has been selected by HM Revenue and Customs (HMRC) for what the company described as a landmark 10-year programme focused on sovereign data infrastructure and AI-enabled decision making across the tax authority.

The announcement comes as HMRC rapidly expands its use of AI and automation technologies across the department. Last month, it confirmed it was rolling out 29,000 Microsoft Copilot licences as part of a broader push to embed AI into day-to-day operations, with HMRC’s chief AI officer James Mitton outlining ambitions to make HMRC the “world’s most AI-enabled tax authority”.
Quantexa said its deal with HMRC represents one of the UK public sector’s largest decision intelligence implementations to date, moving from productivity tooling towards large-scale AI-enabled data infrastructure and operational intelligence.
The company said its technology will help HMRC connect siloed datasets, uncover hidden relationships between entities, and support more informed decision making across areas including fraud detection, compliance and customer services.
“Quantexa is at the forefront of British innovation, developing the technologies we need to power next generation digital public services that are smarter and more secure,” said AI Minister Kanishka Narayan. “This partnership will help HMRC to deploy AI safely at scale, bringing its data together to support better decisions, improve efficiency and deliver a stronger service for taxpayers.
“By using advanced data tools to identify risks and automate routine tasks, HMRC can focus its efforts where they matter most. This is a practical step that paves the way for a better service for taxpayers.”
Sovereign AI
The emphasis on “sovereign” AI infrastructure reflects growing concern across government around data governance, security and operational control as departments scale AI adoption.
Public sector organisations are increasingly being asked to balance rapid AI deployment with concerns around transparency, trust, accountability and data protection, particularly when dealing with highly sensitive citizen information.
If you liked this content…
Quantexa’s systems will reportedly combine HMRC data with external sources to help identify fraud and tax return errors more quickly, while also supporting customer service operations. The company said automated decisions would continue to involve human oversight.
The deal also lands amid mounting scrutiny around HMRC customer service performance and operational pressures. Complaints about HMRC rose to more than 93,000 in 2024-25, with response times among the most common concerns.
Think Digital Partners’ research and events have found that senior public sector leaders are now less focused on experimenting with AI tools and more concerned with how to operationalise AI securely at scale, integrate fragmented data environments, and establish governance around emerging technologies.
That shift is becoming particularly visible in central government departments managing large, complex data estates and citizen-facing services.
Ceri Morgan, HM Trade Commissioner for Europe, said the HMRC deployment also demonstrated the growing international relevance of UK-developed AI and data technologies.
“This announcement underscores the strength of the UK’s advanced data and AI ecosystem, and its relevance to governments around the world,” she said.
“Quantexa’s work with HMRC shows how UK-developed technology can deliver impact at national scale – supporting more efficient public services, safeguarding revenues, and enhancing the taxpayer experience. It’s exactly the kind of innovation we are keen to share with international partners.”
Founded in London in 2016, Quantexa specialises in “decision intelligence” technology, combining AI, analytics and contextual data modelling to help organisations identify risk, fraud and operational inefficiencies across large datasets. The company has grown rapidly in recent years, surpassing $100 million in annual recurring revenue and securing a valuation of $2.6 billion following a recent funding round.








