Over 200 Groups Demand a Halt to AI in Military Kill Chains

A Coalition Sounds the Alarm

More than 200 civil society organisations, led by Access Now and Amnesty International, have issued a joint statement demanding that governments and technology companies immediately halt the use of artificial intelligence in military targeting. Published on 14 June as AI-driven operations continue across multiple conflict zones, the statement warns that AI-accelerated warfare is "rubber-stamping killing at unprecedented speed and scale" and calls on states and companies to "cease the provision and use of technologies which undermine adherence to international law in armed conflicts."

"AI has become the new gunpowder, and tech companies are the new arms dealers," said Marwa Fatafta, MENA Policy and Advocacy Director at Access Now. "From Gaza to Lebanon to Iran, the lived reality of violence unleashed through AI-driven kill chains shatters any claim that these systems can be deployed responsibly."

Claude Inside the Kill Chain

The call follows reporting by The Washington Post that the Palantir Maven Smart System, integrated with Anthropic's Claude, was used during the US-led Operation Epic Fury against Iran to identify and prioritise more than 1,000 targets within 24 hours. Defense One reported that Claude acts as a "reasoning engine" within the platform, synchronising modules and simplifying tasks for human operators. The coalition argues that using AI for target generation and prioritisation pushes militaries into a form of warfare that erodes the foundational principles of international humanitarian law.

Safeguards Lag Behind Deployment

Human Rights Watch, a co-founder of the Stop Killer Robots campaign, says AI adoption in warfare is outpacing the law. In March it criticised the Pentagon's rejection of Anthropic's ethical red lines as "a clear sign that the Pentagon is unlikely to uphold meaningful safeguards." The Trump administration moved the other way on 4 June, signing a National Security Presidential Memorandum directing accelerated AI deployment for "warfighters and intelligence professionals" and tasking Defence Secretary Pete Hegseth with revising weapons-autonomy directives within 90 days. A March editorial in Nature argued that AI in warfare should not be permitted until specific governing rules exist, noting that no international laws explicitly mention AI in armed conflict.

What the Coalition Wants

The statement makes three demands: that states end the use of AI tools in military targeting, that they provide transparency on how AI is currently deployed in hostilities, and that companies ensure their systems do not contribute to violations of international law. "AI companies cannot proclaim to be committed to benefitting all of humanity while simultaneously raking in the profits from lucrative military contracts," said Daniel Leufer of Access Now. "We need the hollow ethical posturing to stop and real accountability to begin." The intervention lands as diplomats gather in Geneva for UN talks on military AI and autonomous weapons.