New Zealand could take a global lead in controlling the development of 'killer robots' — so why isn't it?

“New Zealand versus the killer robots” might sound like a science fiction B-movie, but that was essentially the focus of an event at parliament earlier this month.

Hosted by Minister of Disarmament and Arms Control Phil Twyford, the “Dialogue on Autonomous Weapons Systems and Human Control^[1]” looked at how New Zealand might take more of an international lead in regulating these highly contentious new technologies.

Twyford warned of the danger of warfare “delegated to machines”. He referred to a recent survey^[2] showing widespread public opposition to the deployment of autonomous weapons in war and strong support for government action to ban or limit their development and use.

The prospect of New Zealand’s leadership has been warmly received by activists and campaigners involved in the “killer robots” debate.

Human Rights Watch’s Mary Wareham has argued^[3] New Zealand leadership could act as “a total catalyst for action”, while the Campaign to Stop Killer Robots listed Twyford’s commitment as one of the “key actions and achievements^[4]” of its campaign to date.

Yet New Zealand has not joined the 30 states that have formally called for a ban on autonomous weapons^[5], and Twyford’s statements have tended to waver between bullish and reserved. During the event at parliament he acknowledged the clear ethical problems with autonomous weapons, but also the complexity of making policy.

Sensitivity to military allies

If the mood of the people and government of New Zealand is strongly behind regulation, what makes the issue so difficult?

The short answer is politics and economics. A major obstacle^[6] for Twyford is allowing the New Zealand Defence Force to work with allies and partners.

Read more: Lethal autonomous weapons and World War III: it's not too late to stop the rise of 'killer robots'^[7]

Both the US and Australia are heavily invested in pursuing cutting-edge military technologies, including robotics, artificial intelligence and autonomy. A key pillar of their strategy is building systems that allow more coordination^[8] on the battlefield.

Leading a movement to have these systems regulated or banned could see New Zealand’s military shut out of joint exercises where such technologies are being trialled or used.

Given the political pressure^[9] to take a stronger stand against China, it seems unlikely New Zealand’s Foreign Affairs and Trade or Defence ministries will want to risk further discord with key defence partners.

Protecting high-tech industry

The second hurdle lies in the economic promise of technologies developed in New Zealand that could potentially be used in autonomous weapons programmes elsewhere.

Many leading engineers and technologists have advocated for the regulation or banning^[10] of autonomous weapons, but others are attracted by the potential rewards of military-related projects.

Read more: Killer robots, free will and the illusion of control^[11]

These tensions have already surfaced^[12] in the debate about US military payloads being launched from New Zealand by US-owned aerospace company Rocket Lab.

Autonomous weapons could well see similar questions raised about other technologies developed by New Zealand companies or researchers — most obviously in the fields of computer vision, robotics and swarm intelligence — that could be used in military systems.

Regulating autonomous weapons without also inhibiting potentially lucrative AI and robotics research and development remains a challenge.

Public opinion not enough

The hope that regulation of autonomous weapons could represent another “anti-nuclear moment” in New Zealand’s disarmament and foreign policy history therefore seems premature.

While it’s clear there is support for some form of regulation, there’s little evidence^[13] at this stage to suggest public opinion will sway the government’s current conservative and watchful position.

Read more: AI has already been weaponised – and it shows why we should ban 'killer robots'^[14]

So, what should be done? In the absence of international agreement, New Zealand could press ahead with its own domestic legislation to regulate these technologies, as proposed in a petition^[15] from local Campaign to Stop Killer Robots coordinator Edwina Hughes.

This has the potential to expose a lack of serious commitment to principle in the government’s position, but it would still come up against the political and economic interests opposed to action on autonomous weapons.

Acknowledging those political and economic obstacles is a critical first step for meaningful public debate.

Read more: Never mind killer robots – even the good ones are scarily unpredictable^[16]

Engagement and transparency the key

In the near term, a stocktaking exercise should be undertaken to understand what research and development is being carried out in New Zealand universities and companies.

Efforts should also be made to understand which autonomous technologies are likely to be developed and possibly deployed in the coming years by New Zealand’s major defence partners, particularly Australia and the US.

Serious, sustained dialogue with commercial interests and defence partners is a necessary precondition for the advancement of Twyford’s agenda. While there is some evidence^[17] this work is underway, it needs greater transparency to ensure public understanding of what’s at stake.

Without that, New Zealand will probably struggle to take an international leadership role on this critical issue.

References

1. ^{^} Dialogue on Autonomous Weapons Systems and Human Control (www.beehive.govt.nz)
2. ^{^} recent survey (www.converge.org.nz)
3. ^{^} has argued (www.newsroom.co.nz)
4. ^{^} key actions and achievements (www.stopkillerrobots.org)
5. ^{^} ban on autonomous weapons (www.pgaction.org)
6. ^{^} major obstacle (www.newsroom.co.nz)
7. ^{^} Lethal autonomous weapons and World War III: it's not too late to stop the rise of 'killer robots' (theconversation.com)
8. ^{^} allow more coordination (sldinfo.com)
9. ^{^} political pressure (www.rnz.co.nz)
10. ^{^} regulation or banning (futureoflife.org)
11. ^{^} Killer robots, free will and the illusion of control (theconversation.com)
12. ^{^} already surfaced (www.newshub.co.nz)
13. ^{^} little evidence (www.scoop.co.nz)
14. ^{^} AI has already been weaponised – and it shows why we should ban 'killer robots' (theconversation.com)
15. ^{^} petition (www.parliament.nz)
16. ^{^} Never mind killer robots – even the good ones are scarily unpredictable (theconversation.com)
17. ^{^} some evidence (www.converge.org.nz)