Artificial intelligence has moved from a supporting intelligence role into direct combat operations at a speed that has outstripped the rules meant to govern it, according to military officials, tech executives, weapon makers, and legal experts.
Ukrainian drone startup Vyriy CEO Oleksiy Babenko told an AI arms-makers gathering in Kyiv that supercharging weapons with AI opens a Pandora’s box of killer robots but that the alternative is starker. “Either robots will kill us in 50 years, or the Russians will kill us in a year,” he said.
Former Google CEO Eric Schmidt, who now invests in military-drone companies, said on stage at a recent expo that “future combat will be largely robotic. It will be automated.” He added that it “will be controlled by the laws of war.”
But a central question remains unanswered: whether the last century’s rules can handle warfare’s next era.
Retired U.S. Air Force Lt. Gen. Jack Shanahan, who led the Pentagon’s first major AI application, Project Maven, said “the combination of AI and autonomous weapon systems demands an entirely new approach to risk analysis, risk mitigation, and risk acceptance.”
AI entered combat in 2017 when Project Maven used algorithms to identify ISIS bombers attacking U.S. forces in Iraq. Around 2023, drone makers supplying Ukrainian forces began developing AI to lock onto targets. By 2024, U.S. commanders said they were selecting targets at more than 10 times the tempo of the Iraq war. In the Ukrainian National Guard’s Khartia Corps, automation has tripled the pace of missions, according to the unit’s top drone engineer.
Defense Department Chief Digital and AI Officer Cameron Stanley pushed back against concerns that automation is advancing too fast. “You always have the human who will analyze the situation,” he said at the AI+Expo in Washington. “The most dangerous course of action right now is to stand still and to remain in a human-driven world.”
At the same expo, International Committee of the Red Cross legal adviser Noa Schreuer staffed a Red Cross stand with a sign reading “Humanity in War,” displaying a page from the Geneva Convention and a mock traffic sign for drone operators reading: “Don’t Outsource Your Authority / Maintain human control and judgment.”
“Would an autonomous drone abort a mission on its own, for example if a child enters the target area?” Schreuer asked.
The stakes of that question have been driven home during the current conflict with Iran, according to the report. Early this year, as Pentagon officials said new technology allowed them to identify and hit targets faster, Iranian authorities accused the U.S. of striking a school, killing more than 160 people, many of them children. The Pentagon is investigating whether U.S. forces hit the school near an Iranian military compound and whether AI was involved.
The question of human-machine interaction also vexes Roy Lindelauf, a former Royal Netherlands Air Force Apache attack helicopter pilot who now teaches data science at Tilburg University and the Netherlands Defense Academy. He pointed to “automation bias” — the tendency of people to trust what computers tell them — as a growing concern.
“Even if AI is only a tool, how the human mind works should be taken into account to address biases,” Lindelauf said.
Estonian defense adviser Eva Sula, who works with NATO military leaders, said commanders ask her questions nobody can answer. “A war crime in the digital space?” she said. “No country can prosecute it. They don’t have the laws.”
The Trump administration has sought to address the regulatory gap. President Trump this month issued an executive order on AI in national security that calls for aggressive adoption of the technology but requires it to operate “in accordance with applicable laws, government policies, and guidance.” The order directs the Pentagon to update AI rules adopted only three years ago.
At the same time, the administration has locked horns with AI company Anthropic, whose systems the Pentagon uses. Anthropic sought explicit guarantees that its technology would not be used for mass domestic surveillance or autonomous weapons, sparking administration backlash. MSI previously reported that the Pentagon barred Anthropic from defense contracts in March after talks collapsed last month.
Danylo Tsvok, chief executive of the Ukrainian Defense Ministry’s AI warfare center, A1, said the country accepts battlefield risks linked to AI because “the ethical framework is just now being developed.”
A serviceman on an AI development team told the Journal that for now, “waging war is an art — intuition,” because AI cannot yet respond to unconventional situations.
Pope Leo recently issued an encyclical on AI that built on the work of a Vatican panel of Nobel laureates and tech specialists. The panel’s plea: “AI systems must never be allowed to make life-or-death decisions, especially in military applications.” Panel member Marco Trombetti, CEO of AI-translation company Translated, said the group’s 18 members agreed that “if these things get used for war, they cannot be stopped.”