New Zealand Deploying 50 Troops to US AI Drone Warfare Exercise

New Zealand is set to deploy 50 Defence Force personnel to a United States military exercise centred on AI-assisted drone targeting, raising fresh questions about the extent of the country’s integration into American automated warfare systems.

The New Zealand Defence Force will participate in Project Convergence Capstone 6, a US Army-led joint warfare exercise scheduled for mid-2026 and understood to be held in Arizona. The exercise involves drones equipped with what is described as “autonomous terminal guidance” — technology designed to help identify and strike targets with minimal human delay — as well as augmented reality eyewear developed by defence firm Anduril Industries.

Anduril won a $270 million US Army contract to supply the helmet-and-glasses hardware. The company describes the system as giving soldiers “superhero-like abilities,” with eyewear capable of identifying threats kilometres away and instantly marking them for drone strikes.

The exercise forms part of the broader Combined Joint All-Domain Command and Control initiative, a Pentagon programme that integrates land, sea, air, space and cyber capabilities across allied militaries to enable rapid, AI-assisted targeting. New Zealand has taken part in previous Project Convergence exercises, but this edition escalates the complexity with autonomous targeting systems now central to the programme.

Documents obtained by RNZ under the Official Information Act show that NZDF Legal Services will carry out assessments and develop what the NZDF calls national policy caveats to ensure New Zealand law is complied with during the exercise. Brigadier G A Motley confirmed in the OIA response that when NZDF personnel create or maintain information on systems owned by partner nations, that information “is subject to international agreements.”

The NZDF also stated that it maintains complete control over its own battle management systems during such exercises. Brigadier Motley wrote that “NZDF Legal Services will carry out assessments and develop national policy caveats specific to the activity to ensure that New Zealand law is followed.”

The exercise story arrives alongside intensifying scrutiny of New Zealand’s relationship with Palantir Technologies, the US data and artificial intelligence company founded by New Zealand citizen Peter Thiel. RNZ has separately reported that former Defence Minister Judith Collins met Palantir’s international president Laurence Lee at the Munich Security Conference on 13 February 2026 to discuss upcoming opportunities for New Zealand. Several paragraphs of her briefing document for that meeting were redacted in an OIA release.

Current Defence Minister Chris Penk has since said the NZDF has “no existing plans to use Palantir in the emerging technologies space,” while acknowledging that “the NZDF uses Palantir as an analytics platform to aid with planning.” The Government Communications Security Bureau currently leads New Zealand’s broader Palantir partnership.

The timing of the drone exercise disclosure is significant. Palantir’s Maven AI system — the same system now central to the Combined Joint All-Domain Command and Control initiative — was used to select thousands of targets in US and Israeli airstrikes on Iran in February 2026. Maven is becoming the core targeting infrastructure of the US military and is the backbone of the very system that New Zealand personnel will be training alongside in the upcoming exercise.

A GCSB spokesperson said that “all activities of the NZSIS and GCSB, including the use of technologies, are in accordance with New Zealand law.” However, no details have been made public about what aspects of the GCSB’s Palantir engagement relate to military applications.

Transparency advocates have pointed out that New Zealand’s stated commitment to ensuring a human is in the loop on lethal decisions sits uncomfortably with participation in exercises built around systems designed to accelerate and automate targeting. Previous government defence policy statements had repeatedly emphasised that AI should support rather than replace human judgement in decisions about the use of force. Whether those principles are embedded in the caveats being prepared by NZDF Legal Services has not been made public.

New Zealand’s participation in these exercises has deepened steadily over recent years, broadly in step with the country’s closer alignment with the AUKUS Pillar Two programme, which focuses on advanced technology sharing among Australia, the United Kingdom and the United States. While New Zealand is not a core AUKUS partner, it has been invited to participate in a number of technology-sharing activities, and exercises like this represent practical integration with the same alliance infrastructure.

The question of how much parliamentary scrutiny these commitments receive is becoming a live political issue. New Zealand sends troops to exercises routinely without public announcements or substantive debate. The latest participation became known only through an OIA request. Defence Minister Penk has not made any public statement addressing the autonomous targeting aspects of the exercise.

As New Zealand heads towards a general election, the country’s defence posture — particularly its relationship with the United States and the degree to which it is becoming embedded in AI-driven warfare systems — is emerging as a subject of genuine political contention. The government has emphasised the importance of the alliance with Washington at a time of global instability, but questions about where New Zealand draws its own ethical lines on autonomous weapons remain largely unanswered.

What do you think about New Zealand’s growing involvement in US AI-driven military exercises? Share your thoughts in the comments below.