Workshop report: AI and the future of the public service

Across the workshop, a common thread emerged: AI should provide an opportunity to better connect policy and services to human needs. Participants highlighted that the real benefits lie in improving responsiveness and outcomes for people, not just in deploying new technology for its own sake. While technical solutions are advancing rapidly, many of the underlying problems and opportunities identified are not new —they are consistent with challenges recognised five or even ten years ago.

A key insight was that the fundamental barriers to progress likely do not lie with technology itself, but at the organisational level. Without meaningful changes to structures, incentives, and culture, there is a real risk that governments will be writing similar reports five years from now, with little practical change achieved.

Participants also reflected on how innovation happens across many different settings —inside and outside government —and stressed the importance of partnerships to avoid duplicating existing solutions. They pointed to the need for better mechanisms to fund, adapt, and bring proven innovations to scale within government systems.

Participants also emphasised the critical importance of governance when applying generative AI in government services. They noted concerns about transparency, accountability, and the risks of relying on ‘black box’ systems —particularly the need for clear pathways for reporting and addressing failures when they occur. A risk-based approach was seen as essential to managing these challenges responsibly.