On 1 September 2025, a group of over twenty AI experts in Aotearoa-New Zealand issued an open letter urging the government to step up and regulate artificial intelligence more robustly. The signatories point to a growing lack of trust in AI systems, citing risks ranging from deepfakes and biased decision-making to environmental harm and threats to democratic processes.

Learning from Global Frameworks

Dr Andrew Lensen, Senior Lecturer and AI Programme Director at Victoria University of Wellington—and one of the authors of the letter—highlights regulatory models from abroad that could inspire New Zealand’s approach. He suggests that the EU’s AI Act, which uses a risk-based framework, offers a solid template. Australia’s emerging regulatory efforts could also serve as a useful reference. Importantly, Dr Lensen cautions that New Zealand need not replicate the EU’s extensive and complex system verbatim; rather, it should adapt relevant elements in a way that suits its national context.

The High Cost of Inaction

The consequences of doing nothing are compellingly laid out. According to Dr Lensen, failure to regulate AI could result in both overt and latent harms. Immediate dangers include the proliferation of deepfakes, fraudulent manipulation, and algorithmic bias—especially in sensitive areas like justice or social welfare. But the risks don’t end there: increased CO₂ emissions from energy-hungry AI systems, intellectual property theft, and even erosion of democratic norms are equally concerning. He underscores that “we have an opportunity to decide how we want AI to be used in Aotearoa, and we must seize it.”

A Strategic Imperative

This moment is more than a policy preference—it’s a strategic crossroads. As AI becomes woven into the fabric of society, Aotearoa must determine whether to shape its trajectory or cede control to unchecked technological forces. An effective regulatory framework could help foster innovation while guarding citizens and institutions from harm.

Charting a Path Forward: Key Considerations for Regulation

1. Embrace a Risk-Based Framework

• Adapt, don’t adopt. New Zealand can learn from the EU’s structured risk tiers, tailoring complexity and scope to local needs.

• Proportionality matters. Low-risk applications (e.g., music recommendation algorithms) require lightweight governance; high-risk systems (e.g., AI in healthcare or elections) need more rigorous oversight.

2. Focus on Trust and Transparency

• Accountability mechanisms. Mandate documentation of data sources, model design, and decision logic to enhance explainability.

• Third-party red-teaming. Independent evaluations—similar to how some security systems undergo external audits—could ensure objectivity and detect unintended consequences.

3. Foresee the Indirect Harms

• Regulatory frameworks must account for second-order effects—like the environmental footprint of AI deployments.

• Consider policies that encourage or mandate energy-efficient design and deployment standards.

4. Avoid Regulatory Capture

• Global studies highlight how industry influence can weaken regulatory agendas. To uphold public interest, New Zealand needs strong safeguards—such as independent review boards and diverse stakeholder representation.

5. Engage in Co-Governance

• Increasingly, AI governance is moving away from top-down regulation toward model frameworks that include industry, academia, civil society, and Māori communities. Co-governance ensures policies are equitable and contextually grounded.

Closing Thoughts

AI is transforming economies and societies—this much is clear. What’s uncertain is whether that transformation will benefit all or magnify existing inequalities and systemic vulnerabilities. The call from New Zealand’s AI experts is timely and foundational: responsibly crafted regulation isn’t just a safeguard—it’s a compass for shaping AI’s role in society.

By learning from global frameworks, engaging diverse stakeholders, and preemptively addressing harm, New Zealand has a chance to chart an inclusive, resilient path forward. As Dr Lensen articulates, the choice is ours—and the time to act is now.