Finance Minister Nicola Willis recently announced that the core public service will be cut from about 64,000 positions to 55,000 by 2029, with AI adoption and digitisation cited as key drivers enabling the reduction.
Strategic AI consultancy Ten Past Tomorrow said transforming the public sector using AI is unlikely to succeed unless senior leadership, including ministers and ministry heads, actively adopts and champions AI themselves.
“You can’t expect the public sector to embrace AI if their bosses aren’t themselves AI-skilled and literate. No one can lead a technology they’ve only been briefed on,” Ten Past Tomorrow founder Mark Laurence explained.
“AI transformation doesn’t happen because a minister announces it. It happens because those in leadership have experience of using it powerfully themselves, so they can extrapolate and understand what it would mean to a public service department. That’s not what is currently happening, and that’s what has to change.”
“I will happily give my time to help ministers understand what is genuinely possible with AI,” he said.
“This technology offers us enormous productivity gains, but the people at the top need a deep understanding of it before they can lead others through it.”
Microsoft’s 2026 Work Trend Index, based on trillions of productivity signals and a survey of 20,000 workers globally, found that organisational factors like leadership and manager support account for 67% of AI’s real impact at work, while individual mindset and skills make up the remaining 32%.
“The tools exist, and the case is proven by the private sector,” Laurence said.
“What’s lacking is the leadership confidence to move the public sector from AI as a talking point to AI as embedded practice.”
Laurence said most public sector organisations are lagging in AI adoption because the technology has not yet been strategically integrated into core operations or decision-making processes.
He said the public sector has a major influence on New Zealand’s overall economic productivity and argues that expanding AI use across government agencies is ultimately a national productivity issue.
He added that while efficiency gains from AI are real, they should be seen as a baseline rather than the upper limit of what’s possible.
“Rather than cutting jobs, the more powerful play is to use AI to fundamentally reshape how the public sector works, not just drive small efficiency gains.”
“That’s a much bigger prize, and to get there we need to build leaders who know how to lead through AI.”