The United States is planning broader use of the AI system Maven in military strikes, raising concerns about legal liability and control when AI influences combat decisions.
Ruslan Rashitkhanov, deputy director of the Institute for Information and Media Security at Kutafin Moscow State Law University (MSAL), said using artificial intelligence systems to make combat decisions instead of humans creates a rift in determining responsibility for the use of force.
According to Reuters, the Pentagon plans to permanently integrate the Maven AI system into the U.S. military. The system has already been used in attacks against Iran.
Rashitkanov pointed out that under the current legal framework, AI is considered a tool for information and analytical support, but the legally significant final decision on the use of force must be taken by competent officials. He said that if AI were to effectively replace this human decision-making, it would create a legal gap as there would be no clearly defined decision-making entity and responsibility would be difficult to assign.
This is why Russian law and military doctrine insist that the final decision to use force must be left to humans and no technological system can replace it, he stressed.
He said that when AI outputs directly influence combat decision parameters, the situation goes beyond the internal regulations of information systems and becomes subject to international law governing the means and methods of war, including the 1977 Additional Protocol to the Geneva Conventions.
He explained that the legal evaluation applies to the specific configuration of the algorithm’s use in the use of force, not the algorithm itself. At the same time, there is an obligation to take all possible precautions and adhere to the principles of distinction and proportionality.
He also noted that neither Russian law nor current international law recognizes AI systems as independent entities empowered to make combat decisions.
