How do you build AI workforce readiness in government without losing public trust?
At ADAPT’s 3rd Government Edge in Canberra, ADAPT Principal Research Analyst Peter Hind put that question to leaders from IP Australia, the Department of Home Affairs and the Digital Transformation Agency.
Three themes stood out: involving staff in redesigning their work, setting clear boundaries for responsible use and showing practical results.
Key takeaways:
- Involve staff in mapping workflows and deciding where AI can solve real problems.
- Explain how AI is governed, where human oversight applies and when risks need escalation.
- Plan how to use time saved, and measure whether AI improves work for staff and the public.
Start with the work, then find where AI fits
AI adoption should begin with a real business problem.
Margaret Tregurtha, A/g Director General at IP Australia, describes bringing staff together to map workflows, identify pain points and decide where technology could improve work for employees or customers.
Involving staff this way gives them ownership of the changes and keeps the agency focused on useful applications rather than AI for its own sake.
Simon Cooper, A/g Chief Data and AI Officer and First Assistant Secretary at the Department of Home Affairs, says the department is tailoring its approach to the different needs of its 16,000 staff.
Broad access to Microsoft Copilot lets people experiment, while specialist AI development is relevant to a smaller group. Staff have also built tools from the ground up, including an AI agent that checks work against executive expectations.
Build trust with clear boundaries and visible results
Staff need to understand how AI will be used, how risks are managed and who remains accountable.
Simon Cooper says Home Affairs consults staff when AI is expected to significantly change workflows or roles.
He also describes an assessment framework involving legal and security teams, and senior business leaders.
IP Australia’s Margaret Tregurtha explains that clear risk boundaries help staff know when they can experiment and when they need senior approval.
Lucy Poole, Deputy CEO, Strategy, Planning and Performance at the Digital Transformation Agency, encourages agencies to show staff working examples, measure the results of pilots and share what they learn.
Reusing tested tools and approaches from other agencies can help government move faster while learning from previous experience.
Make time saved count
Efficiency gains don’t need to mean faster deadlines or more work.
Lucy Poole and Margaret Tregurtha say agencies should decide how to use the capacity AI creates, whether that’s more time for customer service, complex work, training or quality assurance.
They also emphasise preserving the knowledge and judgement employees need to check AI-supported work.
AI can change how government work is done, but lasting adoption depends on staff having a voice in that change and seeing its value in practice.