25/05/2026
The New Zealand government is the topic of the week for us AI enthusiasts – but for all the wrong reasons. I don’t write about politics, but I do write about AI and this regrettably demonstrates that the New Zealand government are either panicking or they really don’t understand what they’ve got themselves into, but more likely both.
AI is not an enabler for job cuts – it’s a tool for augmenting human capability, and it demands a better thought-out approach than it’s getting from our policymakers.
Election 2026: AI Adoption on the fly
A few days ago, the government announced plans to shed roughly 13% of the public-service workforce. In a press conference, Finance Minister Nicola Willis insisted that “AI is an enabler” and that new technology would largely justify a reduction of 8,700 jobs.[1][2]
I watched as the numbers of pundits like myself expressed their dismay through LinkedIn and I got to wondering, what did the government do to come to this number.
Searching for some rationale, I dispatched a ChatGPT agent to research the news release and present me with some understanding of the process undertaken to shed over 13% of its current workforce.
The agent took eight minutes. Even on my IPhone 12, 8 mins is a long time.
It found nothing.
Beyond the press release, there is no publicly available modelling, no published cost–benefit analysis, no consulting report, and no visible discovery process to justify the cuts. In fact, the only supporting rationale appears to be that staffing levels are being restored to historic ratios.[1][2][8][9][10]
The announcement admitted as much referring to a specific statement acknowledging the lack of modelling and supporting analysis.
Forgotten realities
Let’s not beat around the bush here: government departments in New Zealand are not known for being anything extraordinary compared with their global peers. Some are better than others, but none of them is a poster child for efficiency. The claim that they will magically wring 13% more productivity out of the current public service landscape by forcing them to adopt AI is nothing more than a pipe dream.
We’ve been given a list of cost-cutting measures and a vague insistence that AI will pick up the slack by:
Reducing back-office functions
Reducing the number of government departments
Reducing inefficiency
Demanding the accelerating uptake of AI
What was provided to support this announcement looked more like what a political aide might write on a bar coaster when he sleeps in and arrives five minutes before lights, camera, action. I would offer more detail if I left a note on the bench to my family.
We deserve better, especially when it comes to AI, a technology that is changing the world around us and making the general population increasingly anxious.
This does nothing but open the government up to critisism because it feels like a pre-election hit out looking for votes, not a well thought out response to soothe the anxiety of the nation. Whose vote, I wouldn’t care to speculate, but it sure won’t be mine. This type of approach toward any kind of commercial or public activity is suicidal.
Forgotten narratives
Interestingly, the government’s own pre-budget speech undermines the narrative that AI will miraculously deliver massive savings.
In her speech to the Business North Harbour group on 19 May 2026, Nicola Willis acknowledged that New Zealand’s public service has been “scared of AI, slow to move to the cloud” and is saddled with a “complex and fragmented set of overlapping IT solutions”.[1][5][6][7]
Willis also admitted that the government is frustrated by the “dangerously slow take up of digital and AI technologies”.[1][5][6][7]
In other words, the very departments they intend to shrink have barely begun to adopt the technology that will soon turn them into efficient government bodies.
I shudder to think how a public service that has been afraid of AI and has struggled to migrate to the cloud will suddenly become so much more productive if AI is forced upon them.
The press announcement offers little substance when it tasks the Chief Digital Officer with embedding AI deployment, without modelling or a roadmap, into all public entities within three years — with the implied expectation of major efficiency gains.[1][3]
A reacquaintance with reality
To understand the true madness of this approach, we need to look at the landscape into which the government proposes implementing a regime of rapid AI adoption.
New Zealand is not just slow at adopting AI in the public sector; we’re lagging behind across the board.
A March 2026 analysis by DataForge paints a sobering picture. New Zealand was the last OECD country to release a national AI strategy, only doing so in July 2025.[11][15]
Because of this delay, businesses and public agencies lacked policy certainty and access to coordinated investment.[4][11][15]
The same analysis notes that 68% of New Zealand SMEs have no plans to evaluate or invest in AI – compared with only 38% in Australia.[11]
Our overall AI adoption rate stands at 37.6%, well behind regional neighbors like Singapore, where adoption surpasses 70%.[11]
In other words, the private sector – with financial incentives and market pressures to innovate – is still figuring out how to use AI effectively.
Expecting government departments, which are traditionally conservative and risk-averse, to achieve more than this in a fraction of the time is fanciful.
Moreover, the DataForge report points out that only 24% of New Zealand’s workforce has received any AI training.[11]
The skills deficit operates at multiple levels – technical experts, managers who understand AI’s potential, and executives who can develop strategy.
Without significant investment in training and hiring, adoption will remain slow, and the hoped-for efficiency gains will remain out of reach.[4][11][13]
A framework without a plan
Another thing worth looking at in light of this announcement is the Public Service AI Framework, released in January 2026.
This framework is meant to guide government agencies as they adopt AI, unfortunately, it comprises little more than a one-page summary of high-level principles.
Commentators note that it is very light on detail, offering almost nothing on privacy, safety or implementation.[12] Agencies are essentially left to figure it out themselves.
The framework may be a starting point, but it is not a plan – and it certainly isn’t a justification for deep staff cuts.
Incidentally, I went searching for funding after a recent release in relation to the AI Activator but was left with more questions than answers. The Crowns showpiece for funding innovation, Callaghan Innovation, is being disestablished and disseminated to the PRO's.
In my opinion the general support for AI adoption by the government is fractured beyond being helpful to anyone other than someone who knows the system and the people that operate it.
My 2c worth for what it’s worth
It makes me more than a little uncomfortable writing an article like this and even more so that I intend to publish it.
In my eyes there’s a time and a place for criticism and it’s rarely in the pages of LinkedIn. But this is one of those times when I feel justified in publishing my indifference towards this announcement.
It’s in this realm of technology development and the introduction of AI that I live and breathe. I haven’t spent my career on the benches of parliament.
My career has been spent in the corridors of enterprise, the boardrooms of New Zealand and Australia’s biggest companies.
I spent the first 10 years of my career, ‘the hard yards’ not pounding the pavement to spread my message, I spent it on the finance floor and in the technology teams. My job was to make sure the engine was maintained and ran smoothly.
Then I left all that to dedicate the remains of my career to AI, whatever that may look like. I’ve since seen what AI looks like on the cutting room floor as I’ve built countless integrations, solutions and apps. So I come to the table with a voice that actually matters, I know what I’m talking about.
Most people have come at this announcement not referencing AI at all, rather from the simple angle of what the hell are they doing? You can’t claim the advantages of cutting jobs and downsizing the government structure in advance. You have to work for it first and let’s be honest, you don’t expect to magically achieve this in three years. Ten maybe, but three is never going to happen.
There’s not really much left to say other than this is clearly more political than it is policy.
Technology generally, not just AI, needs a strategy developed from in-depth analysis. When it comes to government you might hope to get that in three years but more likely ten.
Either way, this announcement looks far from encouraging. The government is way out of its depth when it comes to AI. They should be calling on the experts that live in this space.
Written by: Mike G Robinson
Sources
1. https://www.beehive.govt.nz/speech/pre-budget-speech-business-north-harbour
2. https://www.beehive.govt.nz/release/public-service-be-overhauled
3.https://dns.govt.nz/assets/Digital-government/Digitisation-Government-Programme/NZ-Goverment-Digital-Target-State-February-2026.pdf
4. https://www.mbie.govt.nz/business-and-employment/economic-growth/digital-policy/new-zealands-ai-strategy-investing-with-confidence/addressing-barriers-to-ai-uptake-in-new-zealand
5. https://thespinoff.co.nz/politics/20-05-2026/nicola-willis-cuts-jobs-loves-ai-hates-nz-first
6. https://www.theregister.com/public-sector/2026/05/20/ai-sackings-reach-new-zealand-which-will-use-it-to-eject-14-percent-of-government-staff/
7. https://fudzilla.com/news/ai/61029-kiwi-government-uses-ai-as-an-excuse-to-fire-more-government-workers
8. https://www.rnz.co.nz/news/national/519440/hidden-costs-and-dangers-if-public-servants-are-replaced-by-artificial-intelligence
9. https://www.hcamag.com/nz/specialisation/employment-law/new-zealand-to-axe-8700-jobs-in-massive-public-service-overhaul/536899
10. https://halifax.citynews.ca/2026/05/19/new-zealands-government-plans-to-cut-14-of-public-sector-jobs-to-slash-spending/
11. https://www.dataforge.co.nz/nz-ai-adoption/
12. https://simplyprivacy.co.nz/2026/02/03/responsible-ai-guidance-for-nz-public-sector/
13. https://www.dlapiper.com/en-nz/insights/2026/03/quick-on-the-uptake-new-zealands-new-strategic-approach-to-ai
14. https://www.treasury.govt.nz/publications/an/an-24-06-impact-artificial-intelligence-economic-analysis-html
15. https://www.mbie.govt.nz/business-and-employment/economic-growth/digital-policy/new-zealands-ai-strategy-investing-with-confidence
Mike develops deterministic, event-driven AI systems focused on workflow automation, operational intelligence, and enterprise process orchestration. With a background in Accounting, IT Risk Management, SAP FI, and enterprise transformation across Australia and New Zealand, his work centres on practical AI implementation that improves decision-making, governance, and operational efficiency.
Linkedin: https://linkedin.com/in/mgrnz
See my Linkedin Company page: https://www.linkedin.com/company/mgrnz
See my website: https://mgrnz.com
Get $1,000 in AI build credits with hyperagent.com: https://hyperagent.com/refer/6A3RMHNQ
