The Department for Work and Pensions recorded 128,779 completions of artificial intelligence training across 12 programmes during the year to 30 April 2026, according to figures released under the Freedom of Information Act.
The data, obtained and analysed by Parliament Street, covers AI learning delivered through the Cabinet Office as part of cross-government work to increase digital capability.
The largest volumes came from the two-part AI for All pathway. Part One recorded 73,045 completions and Part Two 53,001. Senior Civil Service training accounted for a further 1,620 completions, while nine more specialised programmes recorded 1,113 between them.
DWP reported spending £112,995 on AI-related learning during the 12-month period. The department administers services including welfare payments, pensions and employment support, giving its workforce access to large volumes of personal and operational information.
The figures show how widely introductory AI training has been distributed across the organisation. Most completions were concentrated in the foundational AI for All programme rather than the smaller specialist courses.
That distribution reflects the skills challenge facing large employers. Advanced AI development may involve relatively small technical teams, while thousands of other staff need enough understanding to use automated tools appropriately within existing jobs.
DWP’s activity sits alongside the government’s wider AI Skills Boost programme, which aims to give 10 million UK workers practical AI skills by 2030. Free courses have been opened to adults across the country through partnerships involving government, technology companies and other organisations.
Government figures published earlier this year showed more than one million free AI courses had been delivered through industry partnerships before the target was expanded. Ministers later reported that the total had exceeded 1.7 million courses since June 2025.
Use of AI inside public services carries additional responsibilities because staff may work with sensitive financial, identity and personal data. Basic literacy therefore needs to cover more than operating generative tools.
Employees also need to understand when information can be entered into a system, how automated output should be checked and when human judgement must remain part of a decision. Those controls become more important where errors can affect payments, eligibility or access to public services.
The DWP figures measure course completions rather than the effect of the training. Evidence on productivity, service performance, error rates and safe deployment will provide a clearer indication of whether the scale of participation produces measurable operational improvements.
You must be logged in to post a comment.