HMRC is using artificial intelligence to analyse Facebook, Instagram and other social media activity as part of its efforts to identify serious tax fraud cases in the UK.
The technology, the department confirmed, will only be applied in criminal investigations and under legal oversight.
Officials maintain that the AI tools are used solely in high-level cases, supported by robust safeguards, and that human decision-makers retain the final say. However, critics warn that the move risks sliding towards mass surveillance, potentially leading to wrongful accusations of innocent taxpayers.
Concerns over the use of AI in government processes have grown in the wake of the Post Office Horizon scandal, which saw technology failures cause devastating consequences for thousands.
The AI monitoring system works alongside HMRC’s long-standing Connect programme, a powerful data analytics network capable of processing billions of records from banks, property transactions and other databases to detect potential tax evasion. Chancellor Rachel Reeves has tasked officials with recovering £7 billion from the UK’s estimated £47 billion tax gap.
As part of this drive, HMRC is trialling AI-powered assistants to flag suspicious tax returns and issue warnings that could later be presented in court. The department is also in talks with around a dozen technology companies to explore further uses of AI in recovering unpaid taxes, including funds linked to offshore accounts.
The Department for Work and Pensions has similarly introduced AI tools for administrative efficiency, with 20,000 staff using them to prepare documents and summarise meetings.
Campaigners have expressed concern that AI systems could begin making automated decisions, despite HMRC’s assurances that a human is always involved in the final process. Recent legal proceedings have even compelled HMRC to disclose whether AI is already influencing research and development tax credit claims, following a Freedom of Information request by tax specialists.
HMRC has stressed that AI-supported social media monitoring is strictly limited to criminal cases, aimed at targeting serious fraud and enabling staff to focus on assisting taxpayers while bringing in more revenue for public services.
