Whitehall’s top official orders review of AI in the civil service

TL;DR:

  • New Cabinet Secretary Dame Antonia Romeo will lead a review of the “organisation, performance and transformation” of the permanent civil service.
  • AI and technology’s impact sits at the centre of the review, which reports in early 2027.
  • Full terms of reference, drawing on a panel of expert advisers, will be published this summer.

The UK’s most senior official, Cabinet Secretary Dame Antonia Romeo, is to lead a review of the permanent civil service that places the impact of AI and technology at its core, PublicTechnology reports. The work will assess the size, shape and structure of the civil service, its productivity, and the government’s ability to attract and retain talent, before reporting in early 2027.

Reform with AI at the centre

Romeo has set out “reforming the civil service so that it is recognised for excellence in delivery, innovation and improved productivity” as a central objective. The review will culminate in a vision document with recommendations on accountability, capability and safeguarding impartiality, alongside a refresh of the civil service code aimed at rebuilding trust with ministers, parliament and the public. A panel of advisers from academia, the private and public sectors and civil society will inform it, with full terms of reference due this summer.

The review lands amid a steady drip of central-government AI activity, from the Government Digital Service’s insistence that good data, not tools, underpins adoption to the national police AI centre set up to cut paperwork. It is being run in parallel with an update to the Cabinet Manual, which Sir Keir Starmer said this month had become “significantly out of date”.

Looking forward

For a workforce of nearly half a million, the review signals that AI is no longer a side project but a question of organisational design — how many people Whitehall needs, doing what, and with which tools. The framing is deliberately broad, but the test will be whether it produces concrete commitments on retraining and recruitment rather than another aspirational vision. With reporting slated for early 2027, departments already piloting AI will be watching whether the centre sets standards they must follow or simply encourages them to keep experimenting.

Share this article