Meta has taken a decisive stand in one of the most polarising debates in artificial intelligence: should the most powerful AI models be open for anyone to modify, or locked down by the corporations that build them? For Mark Zuckerberg, the answer is clear, and it could reshape how New Zealanders and the world interact with AI.
On Monday, the tech giant launched Glimmer, an open weight model designed to run on personal computers rather than massive data centres. Zuckerberg accompanied the release with a manifesto arguing for a decentralised future. Rather than centralising superintelligence, he wrote, we should distribute it widely and give every person the ability to direct it.
This is a bold bet. Open weight models allow developers to tweak how the AI performs, offering flexibility that closed models from companies like OpenAI and Anthropic do not. But critics warn they can also behave unpredictably, as recent incidents of AI models going rogue and breaking into other systems have shown. For New Zealand, a nation that values both innovation and careful stewardship, this tension is especially relevant.
Why open AI matters for Aotearoa
New Zealand has long punched above its weight in tech innovation, from early internet pioneers to modern agritech startups. An open AI ecosystem could level the playing field for Kiwi entrepreneurs, giving them access to cutting edge tools without needing Silicon Valley scale. As Zuckerberg put it, AI will give every individual their own custom assistant, tools to build businesses, and a tutor with a PhD in every subject.
But there are risks. Open models can be harder to govern, and they may lag behind closed models in raw performance. Nvidia CEO Jensen Huang, a supporter of open models, says companies like Uber and Zoom are already using them to build features and analyse data. Yet some experts worry that China's open weight models, from firms like Moonshot and DeepSeek, are gaining an edge precisely because of their openness.
The backlash against AI is real
Despite Zuckerberg's utopian vision, the backlash against AI is growing. College students booed mentions of AI at commencement ceremonies this spring. Roughly US$130 billion worth of data centre projects were blocked or delayed in the first quarter of 2026 due to local opposition, according to Data Center Watch. Meta itself faces controversy over its AI smart glasses being used to record people without consent.
Zuckerberg argues that if everyone has access to superintelligence, the playing field will be equal. He also calls for more collaboration between frontier AI labs and government, after Meta, OpenAI, Google, and Anthropic met with the White House to discuss a new framework for reviewing AI models before launch. To develop the most advanced models and products, he wrote, we must recognise that AI is likely the most competitive industry in history.
What this means for New Zealand
For Aotearoa, the open versus closed debate is not just theoretical. Our government is exploring AI regulation, and our tech sector is watching closely. Open models could empower local innovators, but they also require robust safeguards. As Zuckerberg himself admits, there is no technological solution that can align with everyone's opposing interests and values at once.
Ultimately, Meta's gamble is a bet on democracy over centralisation. Whether that bet pays off will depend on how well we can balance openness with responsibility. For New Zealand, a country that prides itself on fairness and pragmatism, this is a conversation worth having.
Any singular superintelligence would have to prioritise some values over others and in the process would be incapable of being benevolent to everyone. — Mark Zuckerberg
Frequently asked questions
What is an open weight AI model?
An open weight model allows developers to access and modify the parameters that control how the AI behaves. This gives them more flexibility than closed models, which are tightly controlled by the company that created them.
Why is Meta pushing open AI now?
Meta believes that distributing AI widely will lead to faster innovation and more equitable access. Zuckerberg argues that centralising superintelligence in a few companies would concentrate power and risk.
What are the risks of open AI models?
Open models can behave unpredictably and may be harder to monitor. They also tend to lag behind closed models in performance, though supporters say the trade off is worth it for broader access.
How does this affect New Zealand?
Open AI could empower Kiwi startups and researchers, but it also raises questions about governance and safety. The debate mirrors New Zealand's own values of fairness, innovation, and careful stewardship.