New Zealand’s AI regulation push could stifle our tech sector before it gets started

Parliament’s proposed AI oversight framework promises to protect Kiwis from algorithmic bias and data misuse, but the compliance costs and regulatory uncertainty could drive our best tech talent across the Tasman before New Zealand’s AI sector finds its feet.

The government’s latest consultation on artificial intelligence regulation landed with a thud in Wellington last week, promising comprehensive oversight of AI systems used in everything from hiring decisions to loan approvals. On paper, it sounds sensible enough – who doesn’t want protection from biased algorithms deciding your mortgage application? But scratch beneath the surface and you’ll find a regulatory approach that could kill New Zealand’s nascent AI sector in its crib.

The proposed framework would require companies using AI in “high-risk” applications to conduct extensive audits, maintain detailed documentation, and submit to regular government reviews. For multinationals like Google or Microsoft, these requirements are just another compliance box to tick. For a Auckland startup trying to build the next breakthrough in agricultural AI or a Wellington fintech using machine learning to detect fraud, it’s a potential death sentence.

Take Christchurch-based AgriTech firm GrowthMetrics, which recently developed an AI system that can predict crop yields with 94% accuracy. Under the proposed rules, their technology would likely fall under “high-risk” classification because it influences agricultural investment decisions. The compliance costs alone could consume half their current runway, forcing them to choose between burning cash on regulatory paperwork or relocating to Australia where the regulatory environment remains more startup-friendly.

This isn’t just theoretical hand-wringing. We’ve seen this movie before with New Zealand’s financial services sector. The Financial Markets Conduct Act of 2013, while well-intentioned, created such onerous compliance requirements that dozens of small investment firms simply shut down rather than deal with the regulatory burden. The survivors were mostly large Australian banks that could absorb the costs. The result? Less competition, higher fees, and a financial sector dominated by foreign players.

According to the Productivity Commission, New Zealand already lags behind Australia and other OECD countries in digital technology adoption and innovation. The last thing we need is a regulatory framework that widens this gap further. Their research shows that regulatory uncertainty is one of the primary barriers preventing New Zealand businesses from investing in new technologies.

The government argues that light-touch regulation will expose Kiwis to algorithmic discrimination and privacy breaches. Fair enough – nobody wants AI systems making unfair decisions about employment, credit, or healthcare. But there’s a world of difference between targeted regulation that addresses genuine risks and a blanket approach that treats every AI application as potentially dangerous.

Singapore offers a better model. Their Model AI Governance Framework provides clear guidelines for high-risk AI applications while maintaining flexibility for innovation. Companies get certainty about expectations without drowning in compliance paperwork. The result? Singapore has become a regional hub for AI development, attracting both startups and established tech giants.

Meanwhile, our proposed approach reads like it was drafted by lawyers who’ve never built a machine learning model. The definition of “high-risk AI” is so broad it could capture everything from Netflix’s recommendation algorithm to a small retailer’s inventory management system. The compliance requirements are so detailed they’d make a tax accountant weep.

Even more concerning is the timing. New Zealand’s tech sector is finally showing signs of life after years of brain drain to Silicon Valley and Sydney. Companies like Rocket Lab have proven that world-class tech companies can emerge from New Zealand. AI represents our next big opportunity – we have strong universities, a reputation for ethical business practices, and unique challenges in agriculture and environmental management that AI could help solve.

But regulatory overreach could snuff out this opportunity before it develops. Young AI companies need to move fast and iterate quickly. They can’t afford to spend months navigating regulatory approval processes or hire compliance officers before they’ve even validated their business model. The proposed framework would essentially require startups to operate like established corporations from day one.

The government should take a page from the European Union’s AI Act, which uses a risk-based approach but includes specific exemptions for research and development activities. This allows innovation to flourish while still protecting consumers from genuinely harmful applications.

Instead of rushing to regulate everything that contains the letters A and I, Parliament should focus on targeted rules for the genuinely problematic use cases – facial recognition in public spaces, AI-powered hiring systems, automated credit decisions. These applications clearly impact people’s lives and deserve oversight. But requiring the same level of scrutiny for agricultural optimization software or customer service chatbots is regulatory overkill.

New Zealand has a choice to make. We can either position ourselves as a smart, innovation-friendly jurisdiction that attracts AI talent and investment, or we can regulate ourselves into irrelevance while watching our best entrepreneurs decamp to more welcoming shores. Based on the current consultation document, we’re heading firmly toward the latter.

The consultation period closes next month. If you care about New Zealand’s tech future, now’s the time to speak up. Because once we’ve regulated our AI sector out of existence, there won’t be much left to consult about.