evidoria

← Back to browse

Good practice

Australia's Whole-of-Government Microsoft 365 Copilot Trial

Australia · Canberra · See the Australia profile

Australia's DTA ran the world's largest whole-of-government Copilot trial — 5,000+ staff, almost 60 agencies — with an independent Nous Group evaluation finding up to an hour a day saved and 77% satisfaction, while openly flagging job-displacement and de-skilling risks.

5,000+
Trial participants (Jan-Jun 2024)
~60
Agencies involved (Jan-Jun 2024)
69 %
Participants reporting faster task completion
61 %
Participants reporting improved work quality
77 %
Participants satisfied with the tool
86 %
Participants who wanted to keep using Copilot after the trial
~40 %
Participants who reallocated saved time to higher-value work

Details

Maturity
Pilot
Promoter
Digital Transformation Agency (DTA), Australian Government, with independent evaluation by Nous Group
Period
November 2023 announced; trial ran January–June 2024
Keywords
generative AI, digital government, whole-of-government productivity, public sector workforce, AI evaluation

Context

Announced in November 2023, the Australian Government's Digital Transformation Agency (DTA) ran a six-month trial of Microsoft 365 Copilot from January to June 2024, giving over 5,000 Australian Public Service staff across almost 60 agencies access to the generative-AI tool — the largest whole-of-government AI trial conducted anywhere at the time. Nous Group carried out an independent evaluation drawing on input from more than 2,000 trial participants.

Results

The evaluation found participants saved up to an hour a day, mainly on summarising documents, drafting first versions and searching for information; 69% reported faster task completion and 61% reported improved work quality. One in three participants used Copilot daily, 77% were satisfied with the tool, and 86% wanted to keep using it after the trial ended. Around 40% of participants reallocated saved time to higher-value work such as mentoring and stakeholder engagement.

Conclusions

The published evaluation was notably candid about limitations: staff reported stigma around visibly using AI, discomfort with meeting transcription, unpredictable outputs that required verification, and concerns — particularly among women, junior staff and administrative roles — about de-skilling and job displacement, as well as risk from reliance on a single technology vendor. The DTA has since used the findings to inform agency-specific training and a broader rollout of generative AI across the Australian Public Service.

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful