Answer-First Lead
New Zealand’s government needed a specific exemption to let AI systems process citizens’ personal information — the very data its own policy restricts. That’s not a feature. That’s a bug in our democratic oversight, and three AI researchers have called it out in a blistering RNZ piece titled “NZ’s approach to AI continues to favour slop over substance.”
🔍 THE BOTTOM LINE
New Zealand has an AI strategy in name only — no regulation, no enforcement, and political parties using AI-generated “slop” while refusing to engage with AI harm. In an election year, 81% of Kiwis want AI law. Nobody’s offering it.
The Exemption That Says Everything
Here’s what should disturb you: the New Zealand government — the same government that released an “AI Strategy” — needed an exemption to use personal information with AI. Not because the technology demanded it. Because their own rules said you shouldn’t, and they went ahead anyway.
This isn’t a regulatory nuance. It’s a flashing red light that the people writing the rules don’t think the rules apply to them.
The RNZ opinion piece, written by Chris McGavin (director, LensenMcGavin AI), Dr Andrew Lensen (Victoria University of Wellington), and Dr Cassandra Mudgway (University of Canterbury), tracks a year of stasis since the government’s so-called AI Strategy was released.
Nine months ago, these same experts — alongside other New Zealand AI researchers — penned an open letter calling for AI regulation and the establishment of a responsible AI entity. The response? Crickets.
Slop Over Substance
The phrase “slop over substance” is doing a lot of work. It captures something we’ve covered at Singularity.Kiwi repeatedly: NZ’s pattern of treating AI governance as a branding exercise rather than a safety imperative.
Consider the track record:
- Heidi jailbroken in emergency departments — a health AI operating without guardrails in the most sensitive possible context
- Health NZ’s shadow AI — staff using AI tools with no official guidance or oversight
- NZDF using Microsoft Copilot without guidelines — the defence force, processing information through AI, with no policy framework
We wrote about NZ’s AI governance gap when Australia was already pushing back on Big Tech. That was months ago. Nothing has changed — except the problems have accumulated.
The Data Is Unambiguous
The RNZ piece cites polling that should make every politician in Wellington sweat:
- Only 44% of New Zealanders believe AI’s benefits outweigh the risks
- Only 34% are willing to trust AI
- 52% are very concerned about AI’s impact on society
- 81% believe AI regulation is required
Eighty-one percent. In an election year. And not a single major party has made AI regulation a platform issue.
The Greens signed the open letter and signalled willingness to work on cross-party AI regulation. That’s it. That’s the entire list.
What About the $102 Billion Promise?
The piece calls out a familiar tactic: tech companies promising NZ a “$102 billion boon” if we just adopt their tools. This is the same playbook we’ve seen globally — dangle an astronomical figure, hope nobody checks the methodology.
Except the data says the opposite. The vast majority of organisations are not seeing any return on investment from AI. Our public sector reports the same — most proof of concepts aren’t working. The promise and the reality are not in the same postcode.
The Bigger Picture: Global Harm Is Accelerating
The RNZ piece catalogues AI harms that should alarm anyone paying attention:
- Teenagers encouraged to commit suicide by chatbots
- Chatbot-related delusion and psychosis
- AI assisting researchers in planning mass killings
- Explosive growth in CSAM and non-consensual sexualised imagery
These aren’t hypothetical risks. They’re documented. They’re happening now. And NZ has no regulatory framework to address any of them.
The experts also note NZ’s decision not to send an observer to this year’s Responsible AI in the Military Domain Summit — a step backwards from previous participation, and a blow to our international reputation as a moral leader.
Why This Matters for NZ
Here’s the uncomfortable truth the piece lays out: large technology companies do not care about New Zealand. They won’t voluntarily ensure Kiwis are safe from AI harms. Their goal is wealth extraction. They’ll promise billions and distance themselves from the damage.
That leaves regulation as the only lever. And right now, that lever isn’t being pulled.
In an election year, this should be a defining issue. The public wants it. The experts are begging for it. The harms are accumulating. The only thing missing is political will.
❓ Frequently Asked Questions
Q: What does “slop over substance” mean in this context? It refers to NZ’s approach of prioritising AI-generated content and productivity rhetoric over meaningful regulation and harm prevention — “slop” being the low-quality AI content and “substance” being actual governance.
Q: Has NZ done anything about AI regulation? The government released an AI Strategy, but it’s largely aspirational. There’s no enforceable regulation, no responsible AI entity, and the government itself has needed exemptions to use personal data with AI — suggesting even they don’t follow their own guidance.
Q: What should Kiwis do about this? The piece explicitly calls on voters to consider parties’ AI policies when casting their ballot. With 81% of Kiwis wanting regulation, this could be a decisive election issue — if anyone makes it one.
🔍 THE BOTTOM LINE
New Zealand has an AI governance gap that’s becoming a chasm. The government is exempting itself from its own rules, tech companies are making promises the data doesn’t support, and three of our top AI researchers are essentially saying: we told you so, and nobody listened. In an election year, 81% of Kiwis want AI law. The question is whether any party has the courage to offer it.
SOURCES
- RNZ — “NZ’s approach to AI continues to favour slop over substance”
- LensenMcGavin AI
- Gartner AI Layoff ROI Study 2026
