Can New Zealand Afford Not to Control Its AI Destiny?
When we hear "Sovereign AI," it's tempting to think this simply means building our own local language models. For New Zealand, the reality is more fundamental: maintaining control over our AI destiny in an increasingly contested technological landscape.
Data sovereignty means controlling what trains these systems. Most AI models today are trained predominantly on English-language internet content reflecting American or European cultural norms. Our unique perspectives (Te Reo Māori, Pacific languages, bicultural governance, environmental stewardship) risk being marginalised in the systems we increasingly depend on.
Equally critical is where our sensitive data resides and who has access to it. When New Zealand health records, government documents, or commercially sensitive information get processed by foreign AI systems, that data often sits on overseas servers subject to foreign jurisdiction. This creates real risks around privacy breaches, unauthorised access by foreign governments, and loss of control over some of our most valuable and sensitive information assets.
Infrastructure and strategic autonomy address critical dependencies. When we rely entirely on foreign cloud providers for AI capabilities, we face two risks: we're one geopolitical disruption away from losing access entirely, and foreign entities may have visibility into sensitive applications involving defence, intelligence, or critical infrastructure like our power grid and water systems.
For a small, isolated nation, this carries the same risks we've always managed with our electricity grid and telecommunications infrastructure. We need the operational independence in AI that we expect in any other strategic sector.
Regulatory sovereignty lets us set our own rules. We've built distinctive approaches to privacy, indigenous rights, and public sector accountability. This means building systems that reflect values like the principles in Te Tiriti, rather than accepting frameworks imposed by Silicon Valley or Beijing.
Economic sovereignty means building domestic capability rather than being purely consumers of foreign AI services. We're already world-class at agricultural technology, environmental management, and public sector innovation. The opportunity lies in combining our precision agriculture expertise with AI systems trained on our unique farming conditions, or leveraging our environmental data and conservation expertise to create AI solutions that address global climate challenges.
Competitive advantage comes from amplifying what makes us distinctive. We're not trying to build the next ChatGPT. Instead, we should identify where our unique strengths (agricultural innovation, environmental stewardship, bicultural governance, public sector effectiveness) create genuine competitive advantage, then use AI to amplify and export that expertise.
Countries like France (Mistral), the UAE (Falcon models), and various European initiatives pursue sovereign AI because they recognise it as essential infrastructure for economic and national security. Think of it like energy independence or food security.
When other nations face similar agricultural or environmental challenges, they could be buying New Zealand AI solutions built on our decades of specialised knowledge and data, rather than us simply importing generic foreign products.
For New Zealand, the question becomes whether we can afford not to invest in sovereign AI capability, given how central these systems will be to everything from our agricultural and intellectual exports to our public services in the decades ahead.
Join us at 7pm on Thursday 27th November where we'll be lifting the lid on NZ's sovereign AI infrastructure, with special guests Paul Seiler and the team from Catalyst Cloud. Register now:
