AustraliaPublic sector2024 – ongoingDepartment of the Treasury
Microsoft 365 Copilot administrative productivity trial
55% of Treasury post-trial respondents reported at least moderate improvement in preparing meeting minutes with Copilot
Problem
Treasury wanted to assess whether generative AI could improve routine administrative work.
Approach
Treasury ran a 14-week Microsoft 365 Copilot trial from May to August 2024. Staff used it within Microsoft Office applications, with training and requirements to check outputs. The Australian Centre for Evaluation conducted an internal evaluation using surveys and focus groups.
Results
Takeaway
The results reflected voluntary self-reports. The evaluation lacked a comparison group that could establish whether Copilot caused the improvements.
Categories
- Industry
- Public sector
- Business function
- Internal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Company's own claim
Sources
Source summaries
Treasury's internal evaluation reported perceived improvements in administrative tasks. Among post-trial respondents, 55% reported at least moderate improvement in meeting-minutes preparation and 59% in summarisation. These percentages measured perceived improvement, rather than time saved.
Full details are in the original sources.