Gary Corn, a former military attorney who now teaches at American University’s Washington College of Law, argues that artificial intelligence is redefining the battlefield by making deception faster and more sophisticated. He explains that while military deception is long accepted as a legal tactic in war, the introduction of AI-generated lies complicates the assignment of responsibility. "It is an axiom that military deception is a legitimate method of warfare," Corn writes, but he adds that when machines produce the deceptions, tracing legal accountability becomes increasingly murky.
Corn highlights how AI systems are already being used to generate synthetic battlefield signals that mimic real military activity, such as radar returns, acoustic noise, or even fake troop movements. The goal is to confuse enemy sensors and mislead them into targeting wrong locations. However, when these AI-generated deceptions lead to civilian casualties, the question of who is responsible under international humanitarian law becomes urgent. The legal framework struggles to keep pace with how quickly AI can create and deploy such fakes.
The Speed and Limitations of AI
One key challenge Corn identifies is the sheer speed at which AI systems process and generate battlefield data. While this rapid processing offers tactical advantages, it can also overwhelm human operators. He refers to this as "human sensory overload," a situation in which operators may not have enough time to assess, analyze, or override machine-generated recommendations. This is especially concerning given the limitations of current AI models. They can be opaque in how they arrive at conclusions, prone to unpredictable behavior, and often produce outputs that humans struggle to verify.
Corn points to the U.S. Department of Defense’s Project Maven as an example of how AI is being deployed to sift through massive data sets from battlefield sensors. The project aimed to use machine learning to reduce the workload on analysts and improve the speed and accuracy of targeting decisions. The hope was to meet legal obligations by enhancing situational awareness and reducing preventable mistakes. Yet, the very same systems that aim to reduce error can also be used to fabricate data, intentionally mislead, and create new forms of battlefield fog.
Machines That Lie to Machines
The next frontier, according to Corn, is machine-to-machine deception, where AI-generated content is directly fed into enemy systems without human intervention. This is a shift from traditional deception tactics, which rely on misleading human observers. AI can now generate synthetic battlefield data so convincing that it is indistinguishable from real information to other algorithms. The result is a battlefield in which both sides may be reacting to fabricated inputs, making strategic choices based on faulty or false data.
Christian Brose, a strategist at Anduril Industries, has warned that as sensor technologies become more pervasive, the ability to remain undetected on the battlefield is rapidly diminishing. "In a world that is becoming one giant sensor, hiding and penetrating — never easy in warfare — will be far more difficult, if not impossible," he said. AI is the tool that enables forces to blur the lines between real and fabricated, creating a battlefield environment in which nothing can be taken for granted.
Corn notes that the challenge is not just technical but deeply legal. International law governing armed conflict assumes a degree of human control over the conduct of war. But as AI systems become more autonomous, the role of human oversight becomes more ambiguous. If a system generates misleading data that leads to an illegal attack, who is responsible? The machine, its operator, the software’s creator, or the commander who authorized its use? These questions remain unanswered and underscore the need for updated legal frameworks that account for the realities of modern AI warfare.
The broader implication is that as artificial intelligence reshapes the way war is conducted, the law must adapt to ensure that accountability is maintained. Corn warns that without clear legal definitions and lines of responsibility, the benefits of AI could be overshadowed by unintended consequences, especially when civilians become casualties of algorithmic deception.