The United Nations is sounding an alarm about autonomous weapons systems that are already deployed in active conflicts, with artificial intelligence algorithms making targeting decisions that result in civilian deaths. While no country has publicly admitted to deploying fully autonomous killing machines, UN Secretary-General António Guterres and human rights experts warn that the technology is advancing faster than international law can regulate it, creating a dangerous gap between military capability and ethical oversight.

What Exactly Are Autonomous Weapons Doing Right Now?

The weapons in question are primarily unmanned aerial vehicles, commonly known as drones, which are increasingly equipped with artificial intelligence systems that identify targets, select weapons, and present firing decisions to human operators. However, the speed and complexity of these systems mean that human decision-making is becoming more theoretical than practical. According to the UN, algorithms are already responsible for civilian deaths in major conflict zones including the Israel-Palestine conflict, the Russia-Ukraine war, and various civil conflicts in Africa and Asia.

The Pentagon's budget request for next year allocates approximately $55 billion specifically for creating a new unmanned, AI-powered arsenal, signaling that the United States views autonomous weapons as a central component of future military strategy. This investment dwarfs traditional defense spending on items like fighter jets and tanks, which can take years to develop and cost over $100 million per unit.

"Algorithms are choosing bombing targets and are already responsible for killing civilians in some major conflict zones," stated Simon Adams, Professor of Human Rights at Murdoch University and former President and CEO of the Center for Victims of Torture.

Simon Adams, Professor of Human Rights at Murdoch University

Why Is a Global Ban So Difficult to Achieve?

The challenge in regulating autonomous weapons stems from the fact that major military powers, particularly those with permanent seats on the UN Security Council, are increasingly dependent on drone and AI systems for warfare. These nations have little incentive to voluntarily restrict technologies that give them military advantages. Additionally, the definition of what constitutes a "fully autonomous" weapon remains contested; militaries argue that human operators maintain ultimate control even when AI systems perform targeting and weapon selection.

Guterres has called for autonomous weapons to be banned by international law, arguing that some decisions must remain forever human, particularly decisions about taking human life. However, translating this moral imperative into binding international law faces significant obstacles from nations that view these systems as strategically essential.

What Are Experts Proposing as Solutions?

The international community is exploring multiple pathways to regulate autonomous weapons, though experts disagree on which approach would be most effective. Some proposals focus on banning the weapons entirely, while others target specific components or uses. Here are the main regulatory approaches being discussed:

• Complete Autonomous Weapons Ban: Guterres and human rights advocates argue for a comprehensive international treaty similar to the Treaty on the Prohibition of Nuclear Weapons, which would prohibit the development, production, and use of fully autonomous lethal weapons systems across all signatory nations.
• Weaponless Drone Treaty: Some experts propose banning weapons from drones rather than banning autonomous systems entirely, which would allow surveillance and reconnaissance drones to operate while preventing them from being used as killing machines.
• Human-in-the-Loop Requirements: A middle-ground approach would mandate that humans retain meaningful control over targeting and firing decisions, though critics argue this is difficult to enforce and verify in practice.

"Killer robots are a horror that belongs in science fiction. There is nothing more sinister than outsourcing killing and warfighting to emotionless, faceless machines that will select which humans get to live or die," declared Simon Adams.

Simon Adams, Professor of Human Rights at Murdoch University

David Swanson, campaign coordinator for RootsAction, noted that dozens of national governments have already expressed support for banning autonomous weapons, and dozens more have indicated their inclination to support such a ban. This suggests that a treaty could potentially be established among willing nations, following the model of the Treaty on the Prohibition of Nuclear Weapons, with smaller and medium-sized nations potentially serving as initial signatories.

How Could a Treaty Actually Get Implemented?

The most realistic pathway forward, according to experts, involves starting with nations that are willing to defy pressure from major military powers. Smaller and medium-sized countries could establish an initial treaty framework, similar to how the nuclear weapons ban treaty was created, and then work to expand participation over time. This approach would not require consensus from the United States, Russia, or China, which would likely oppose restrictions on their military capabilities.

"We are calling for a treaty to remove all weapons from drones, rather than to ban drones controlled autonomously by AI," explained Nick Mottern, co-coordinator of the Weaponized Drone Ban Treaty Campaign.

Nick Mottern, Co-coordinator of the Weaponized Drone Ban Treaty Campaign

However, Mottern and other experts acknowledge that militaries will likely claim they maintain human control over AI-augmented systems even when the AI identifies targets, selects weapons, and presents all elements of the firing decision to a human operator. This semantic distinction between "human control" and "meaningful human control" represents one of the central challenges in creating enforceable international law on autonomous weapons.

The UN's recent warnings about autonomous weapons come as global conflicts have become more protracted and complex. In 2025 alone, the world faced more than 120 active conflicts, many of which now incorporate AI-powered military technology. Without international agreement on autonomous weapons regulation, experts warn that the technology will continue to proliferate, making future conflicts more unpredictable and potentially more devastating to civilian populations.