DWP finds Copilot saves civil servants 19 minutes a day
Briefly

DWP finds Copilot saves civil servants 19 minutes a day
"Microsoft Copilot saved civil servants 19 minutes daily on routine tasks, according to Department for Work and Pensions (DWP) research comparing users to a control group of non-users. This finding is lower than the 26-minute saving reported by participants in a Government Digital Service research study of 20,000 civil servants published in June 2025, though that research lacked a comparison group."
"The DWP research, examined how the paid licensed version of Copilot performed across eight routine tasks. It saved the most time - 26 minutes daily - when searching for existing information, followed by summarizing information, writing emails, and producing written material. It saved the least time (9 minutes daily) transcribing or summarizing meetings."
"Copilot users primarily redirected time saved toward completing other tasks, improving work quality, and planning. "It allows me to focus less on the mundane stuff. I can spend time doing my actual job," said one participant quoted in the research. Beyond time savings, users reported Copilot improved their work quality, particularly for writing emails - although more for first drafts and suggestions than anything involving human judgement. Users also reported higher job satisfaction, with some calling the service a "comfort blanket" that reduced stress."
DWP research compared Copilot users with a control group of non-users and found an average saving of 19 minutes per day on routine tasks. A June 2025 Government Digital Service survey reported a 26-minute self-reported saving without a comparison group. A separate Department for Business and Trade trial found no overall productivity gains. The DWP evaluation measured licensed Copilot across eight tasks, with largest savings when searching for existing information and smallest savings for transcribing or summarizing meetings. Users mostly redirected saved time to other tasks, reported improved draft quality for emails, and higher job satisfaction. The study surveyed 1,716 users and 2,535 non-users, plus 19 interviews, and applied statistical adjustments for selection bias.
Read at Theregister
Unable to calculate read time
[
|
]