HMRC’s AI Tool for Tax Collection: What It Means in 2026

Business professionals in meeting with laptops discussing strategy

What HMRC Is Actually Doing with AI

HM Revenue and Customs has signed a 10-year, £175 million contract with London-based financial intelligence firm Quantexa to deploy AI systems designed to detect fraud and catch errors across millions of tax filings, alongside an expanded partnership with Microsoft to roll out generative AI across its operations. This is one of the largest government AI procurement deals in UK history and reflects HMRC’s broader push toward automated, data-driven compliance.

AI is not entirely new at HMRC. The department has used machine learning and natural language processing for years, including for compliance targeting, document analysis, debt risk prediction, and customer contact handling, through systems such as the long-running “Connect” data analysis platform launched in 2010. What has changed is the scale and speed generative AI now makes possible.

Why HMRC Is Investing Heavily in AI

The UK’s tax gap, the difference between tax owed and tax actually collected, is estimated at around £36 billion annually. Traditional audit methods struggle to scale against the volume and sophistication of both errors and fraud across millions of annual filings, which is the core problem HMRC’s AI investment is aimed at addressing. Rising complaint volumes and falling compliance productivity have added further pressure to modernise.

Key Elements of HMRC’s AI Strategy

Fraud and Error Detection (Quantexa)

Quantexa’s platform uses contextual decision intelligence to connect data points across large datasets, similar to how an experienced fraud investigator would cross-reference information, including linking shell companies and tracing hidden ownership structures.

Generative AI for Customer Service and Internal Operations

Through its expanded Microsoft partnership, HMRC is rolling out generative AI tools, including Copilot deployments and GOV.UK-wide chat capabilities, aimed at improving productivity, decision-making, and taxpayer-facing service journeys.

Automated Risk Assessment and Nudges

HMRC’s transformation roadmap outlines greater use of AI analytical tools to assess compliance risk and provide automated nudges intended to help taxpayers pay what they owe before formal enquiries become necessary.

Making Tax Digital (MTD)

From April 2026, Making Tax Digital becomes compulsory for sole traders and landlords, requiring digital record-keeping through compatible software as part of HMRC’s wider digitisation of the tax system.

What This Means for Taxpayers and Businesses

Change Practical implication
AI-driven risk assessment Larger businesses, certain industries (construction, hospitality, motor trade), and specific locations may face more targeted scrutiny
Cross-referenced data (Connect and Quantexa) Inconsistencies between returns and other data sources are more likely to be flagged
Mandatory Making Tax Digital (from April 2026) Sole traders and landlords need compatible digital record-keeping software
Generative AI customer service tools Faster, more automated interactions, though human oversight remains part of the process

Governance and Ethical Guidance

Government use of AI, including at HMRC, is governed by the UK’s Data and AI Ethics Framework, updated in December 2025, which sets out principles of fairness, accountability, transparency, privacy, and safety for public sector AI use. Professional guidance to tax advisers also emphasises that AI is a tool to assist, not replace, professional judgment, and that advisers remain responsible for complying with all relevant laws and regulations.

Frequently Asked Questions

Is HMRC using AI to decide who gets investigated?

HMRC uses AI-supported risk assessment as part of deciding which businesses and individuals to examine more closely, based on factors such as business size, industry, and location, though final compliance decisions involve human oversight.

What is the Quantexa contract with HMRC?

Quantexa signed a 10-year, £175 million contract to deploy AI systems that detect fraud and catch errors across tax filings by connecting data points across large datasets to identify hidden patterns.

Does Making Tax Digital use AI?

Making Tax Digital itself is primarily a digital record-keeping requirement, but it feeds structured data into HMRC’s broader systems, which increasingly use AI for risk assessment and error detection.

Will AI replace human tax inspectors at HMRC?

Current HMRC and government guidance frames AI as a tool to support, not replace, human judgment, with AI handling data analysis and risk flagging while trained staff make final compliance decisions.

Final Thoughts

HMRC’s AI investment, spanning fraud detection, generative AI customer service, automated risk assessment, and Making Tax Digital, represents one of the most significant AI-driven transformations in UK government to date. For businesses and self-employed taxpayers, the practical takeaway is that data consistency across returns and other records matters more than ever, as HMRC’s ability to cross-reference information at scale continues to expand.

Related reading: The GenAI Divide: State of AI in Business 2025 and AI Business Solutions.

About the author: The AI Uptrend editorial team covers AI tools, platforms, and industry trends to help readers evaluate new technology with a clear, practical lens.