NZ missing from global statement on keeping AI under human control

Good Ancestors NZ is calling on New Zealand to urgently add its name to a statement by leaders from 20 countries calling for frontier AI models to remain under human control.

The statement was published this week by Finland’s President Alexander Stubb, ahead of the United Nations General Assembly. Australia, Canada, Singapore, Ireland, Germany, Spain, Kenya and South Africa are among the countries that have signed, along with the European Commission.

Good Ancestors NZ Executive Director Dylan Chambers says New Zealand is conspicuous by its absence. The statement remains open to other countries to endorse it, and it is important New Zealand adds its voice to the international debate.

“The real risks of AI are being seen already, whether it’s voice cloning, deepfakes, fraud, or threats to democracy. This move is a positive step to mitigate some of these risks and there is no good reason for New Zealand’s name to be missing.”

The leaders warn that the rapid development of frontier AI models “poses serious risks to safety and security if not appropriately managed”, citing recent cases of capable AI systems “circumventing testing safeguards, exploiting vulnerabilities and gaining unauthorized access to real-world systems”. They call for:

companies to develop transparent safety protocols, including mandatory pre-deployment testing and independent evaluation, with qualified evaluators given sufficient access to assess risks
governments and regional organisations to coordinate common standards and strengthen transparency, including shared reporting of serious safety incidents and ensuring countries across all regions have access to scientific capacity, expertise and trusted evaluation
UN member states to explore creating an international institution able to set standards, enable verification and convene states when capability thresholds are crossed.

Dylan Chambers says adding New Zealand’s name would be a straightforward step that puts the country alongside its international partners.

“We need to stop watching from the sidelines. These countries have come together to say publicly that AI must stay under human control, and that the systems these companies build should be tested and independently evaluated before release.

“New Zealand depends entirely on AI models built offshore, so we have as much at stake in how they are tested and governed as anyone.”

Chambers says the three calls in the statement line up closely with what Good Ancestors NZ has recommended in its policy papers, including requiring leading AI companies to report incidents and risks to the New Zealand government, as they already do in California and the European Union.

He says the first step for the Government is to establish a New Zealand AI Safety Institute, creating domestic capability to monitor risks and advise Ministers. It would also connect New Zealand to the international network of AI safety institutes. A national AI Safety Strategy should follow.

“Signing a statement is the easy part. The harder part is having the people and institutions to act on it, and that is what New Zealand still lacks.”

With the election weeks away, Good Ancestors NZ is calling on every political party to release a detailed AI policy, including stating clearly whether they would add New Zealand’s name to the statement.

“New Zealanders deserve to know what each party would actually do about the risks of increasingly powerful AI, including whether they would put New Zealand alongside other countries calling for stronger safeguards.”

ENDS