UK Government Trial finds Copilot is a timesaver

Civil servants using Microsoft 365 Copilot saved an average of 26 minutes per day during a major UK government trial, according to a new report that could shape the future of AI adoption across the UK public sector.

The three-month experiment, conducted from September to December 2024 by the Government Digital Service (GDS), involved 20,000 government employees across 12 departments and agencies, making it the largest deployment of the AI assistant tool in any organization to date.

The findings, published in June 2025, show that 82% of users would not want to return to their pre-Copilot working conditions, with satisfaction scores averaging 7.7 out of 10 and recommendation scores reaching 8.2 out of 10.

"Trial participants saved an average of 26 minutes a day when using M365 Copilot," the report states. "If this was to be replicated across a full working year, users could save 13 days."

Strong Adoption Across Departments

The trial maintained an 80% adoption rate throughout the experiment period, with participating organizations including major departments such as the Department for Work and Pensions, HM Revenue & Customs, the Home Office, and the Ministry of Justice.

Each organisation committed to at least 1000 licences. This was the largest deployment of M365 Copilot in all organisations, with previous trials and pilots limited to 300 licences at most.

Active usage of M365 Copilot in Microsoft Tools over time represented by adoption rate from October 2024 to December 2024.

Microsoft Teams proved the most popular application for the AI tool, with 71% adoption, while Excel and PowerPoint saw lower uptake at 23% and 24% respectively. Daily usage was centred around communications, with 34% of users employing Copilot daily in Teams and 33% in Outlook.

One user from the Innovation, Digital Development and CoE team described their experience: "I work in IDD/CoE and I've been using M365 Copilot to do most things from communication writing, proofreading, creating presentations, drafting documents, creating images."

The tool demonstrated particular effectiveness in content creation tasks, with document drafting saving users an average of 24 minutes and presentation creation saving 19 minutes. Even smaller, routine tasks like scheduling meetings showed consistent time savings of around 9 minutes.

Benefits Beyond Time Savings

Beyond productivity gains, the report highlighted significant benefits for accessibility and inclusion. Users with disabilities reported that Copilot "significantly aided individuals with disabilities," with testimonials praising its ability to help with dyslexia and dyspraxia.

The tool also showed strong performance in reducing administrative burden, with 85% of users agreeing it provided good value to their organization and 63% saying access to Copilot influenced their job satisfaction.

Challenges and Limitations

However, the experiment also revealed important limitations. The AI assistant struggled with complex, nuanced work requiring human judgment, particularly in policy-focused roles dealing with contrasting opinions or sensitive data.

"M365 Copilot's ability to extract key themes and insights from documents is strong, but it struggles with nuanced or context-heavy data requiring human judgement," one policy participant observed.

Finance users also noted accuracy concerns: "While M365 Copilot efficiently generates initial summaries and reports, it struggles with complex data requiring contextual input and often relies on external data sources without built-in verification."

In sensitive areas, HR professionals expressed particular caution: "In sensitive areas like grievance handling or performance evaluations, there is significant concern about the accuracy of M365 Copilot's outputs, as any errors could lead to reputational risks."

Project delivery specialists emphasized the continued need for human oversight: "For risk mitigation, M365 Copilot can identify trends, but decisions about which risks to escalate or how to approach them still require deep understanding and human judgment."

Future Implications

The report found a clear correlation between users' familiarity with AI tools and the time savings they achieved, suggesting that comprehensive training programs will be essential for successful rollouts.

Professions showing the lowest familiarity and confidence in AI tools saw correspondingly lower benefits, with the report recommending "well thought out change management programmes, including communications, engagement and training."

Read the full report here.