Public-sector AI deployment is outrunning assurance, transparency and evaluation capacity

Humphrey tools are rolling out across Whitehall and 2.4 million NHS chest X-rays a year are AI-assisted. The Public Accounts Committee found in March 2025 that 28 per cent of government IT systems are end-of-life legacy, that around half of civil-service digital roles advertised in 2024 went unfilled, and that 21 of the 72 highest-risk legacy systems lack remediation funding. Only 33 Algorithmic Transparency Recording Standard records had been published despite the standard being mandatory for central government. No independent body audits whether deployed tools work, and the NAO flagged fragmented accountability between DSIT and the Cabinet Office.

Government is the UK's biggest AI adopter and its least assured. A high-profile public-sector AI failure without transparency or audit trails would set adoption back years, and it would corrode trust at the moment institutional capacity matters most.

A statutory ATRS publication duty with enforcement. An independent public-sector AI evaluation and audit function, NAO-linked or on CLTR's three-lines model, would publish before-and-after performance of tools like Consult, and ring-fenced remediation funding would cover the unfunded high-risk legacy systems.

// State-led: Instrument: statutory ATRS publication duty plus NAO-linked audit function and ring-fenced funding.

Government is the biggest AI adopter while its mandatory transparency standard sits barely used and nothing audits whether tools work; one visible failure could set adoption back years.