The government must urgently upskill civil servants to make effective use of AI, MPs have warned.
In a new report published on Wednesday, Parliament’s Public Accounts Committee (PAC) said that gaps in digital advancements among government departments are leading to service backlogs, inefficiencies, and poorer outcomes for citizens.
Sir Geoffrey Clifton-Brown MP, PAC chair, argued: “The government cannot expect civil servants to become magically more productive simply because it purchases AI platforms to run on their computers”.
“To ensure that AI is used safely and effectively to transform services for the citizen, those at the sharp end of deploying it must be actively upskilled in its use.”
The report, Smarter Delivery of Public Services, calls for a clearer digital skills framework across the Operational Delivery Profession (ODP), which represents more than half of the civil service’s 290,000-strong workforce.
It warns that without targeted upskilling, government departments risk relying on “elite crisis teams” as a sticking plaster for bigger gaps in adoption.
‘Do more with less’
“The recommendation to upskill civil servants to make greater use of AI and emerging technology is very helpful, and yet not surprising”, said Gareth Oldale, head of data privacy and cybersecurity at law firm TLT.
“The challenge in the public sector remains to do more with less, whether in law enforcement using AI to tackle cyber crime, or in healthcare diagnosing conditions. The benefits are huge, but so is the skills gap”.
What’s more, Oldale added that competition for top AI talent is “fierce”, with the public sector needing to balance recruitment of tech experts with comprehensive upskilling for existing staff.
“The public sector also has the added challenge of enforcing responsible AI use across industries – so regulators need the same expertise to hold others to account”, he said.
Public sector lag could slow AI rollout
The report follows months of debate over how to modernise the UK’s civil service and integrate AI responsibly.
Recent evaluations by the Department for Business and Trade (DBT) found that pilots of Microsoft’s M365 Copilot tool saved time but did not lead to measurable productivity gains, underlining concerns that tech adoption without foundational data and skills strategies risks falling short.
Meanwhile, research by Pluralsight underscored the issue, revealing that while 95 per cent of UK business leaders view tech upskilling as a strategic priority, half of employees say they lack time for training.
It also found that nearly half of AI projects in 2025 were abandoned midway due to skill shortages.
Toby Hough, vice president of people and culture at HiBob, argued: “If Britain truly wants to become an AI ‘superpower’, we must invest not just in the technology but in the people using it”.
“The real challenge isn’t technological, it’s educational. Give people the tools, training and confidence they need, and AI becomes a catalyst for productivity and growth”.
That progress also depends on strengthening the UK’s National Data Strategy, last updated in 2022, to ensure that government data is interoperable, secure, and usable by AI systems. Without it, the pilot projects risk remaining siloed and unscalable.