HMRC has acknowledged for the first time that it employs artificial intelligence (AI) to monitor taxpayers’ social media accounts as part of criminal investigations into tax fraud. The tax authority stated that AI tools complement traditional checks to analyse online posts, such as those about expensive holidays or large purchases, when these appear inconsistent with a person’s declared income. Officials assert that the technology is used solely in criminal cases, with “robust safeguards in place” and within legal boundaries.
This disclosure arises amid growing concerns in Westminster regarding the expanding role of AI in tax enforcement and fears that its use could become more widespread. Senior Conservative MPs have expressed concerns that reliance on automated tools might lead to errors due to insufficient human oversight. Bob Blackman MP remarked, “If they start taking legal action against individuals based on that, it seems draconian… Without a human check, you can see there’s going to be a problem.”
Sir John Hayes, former security minister and chair of the Common Sense Group of Tory MPs, compared the situation to the Post Office Horizon scandal, stating, “The idea that a machine must always be right is what led to the Post Office scandal. I am a huge AI sceptic.”
The AI monitoring tools function alongside Connect, HMRC’s data analytics system used for routine tax investigations. Introduced over a decade ago, Connect cross-references billions of data points, from bank transactions to property records, to identify potential tax evasion.
Chancellor Rachel Reeves aims to recover £7 billion of the UK’s £47 billion “tax gap,” with HMRC officials recently publishing a strategy that envisions AI being integrated into “everyday” tax processes. The department is testing AI-powered “assistants” to help the public complete tax returns and support compliance officers in reviewing them. If patterns in a return suggest false information, the system could issue a warning that might later serve as evidence if fraud is established.
Concerns regarding AI’s role in decision-making intensified after a tribunal ordered HMRC to disclose by 18 September whether AI was used in assessing claims for research and development tax credits. The ruling followed a Freedom of Information request from tax expert Tom Elsbury, who suggested AI might already determine the outcome of some claims.
Ministers assert that there is always a human “in the loop” for decisions affecting individuals, and HMRC insists humans have the “final say” in enforcement actions. The Department for Work and Pensions has also trialled AI tools, with 20,000 civil servants using the technology to draft documents and summarise meetings. A government source revealed that HMRC has approached around a dozen tech firms for proposals on using AI to help close the £46.8 billion in unpaid tax, much of which is linked to offshore accounts.
An HMRC spokesperson stated, “Use of AI for social media monitoring is restricted to criminal investigations and subject to legal oversight. AI supports our processes but does not replace human decision-making. Greater use of AI will enable our staff to spend less time on admin and more time helping taxpayers, as well as better target fraud and evasion.”
