In a sworn statement within a lawsuit against Elon Musk, a senior CIA official disclosed that the xAI data centers are allegedly polluting Black communities, but more strikingly, revealed that a generative AI model named Grok Gov—a commercial offering—was used to identify 2,000 distinct targets in Iran within just 96 hours. This revelation underscores the growing integration of artificial intelligence into military operations, with Operation Epic Fury reportedly assisted by AI dashboards that presented data to aid decision-making.
AI in Modern Warfare: Capabilities and Concerns
According to the CIA official, these AI models do not make final decisions but serve as powerful tools for identifying potential targets. The speed and efficiency of AI, which can analyze satellite imagery, communications, drones, radar, social media, weather, and multiple data sources simultaneously, are seen as indispensable in modern warfare where decisions are made in milliseconds. However, this reliance raises significant concerns among policymakers and researchers.
An Al Jazeera report highlighted that a US-led military facility in southern Israel, the Civil-Military Coordination Centre, built to manage the Gaza ceasefire in 2025, was using data from private AI tech companies like Dataminr to issue early warnings. This trend of militaries outsourcing critical analysis to private corporations is growing.
Two Major Problems with AI in Warfare
The first problem is that AI algorithms are prone to error. For instance, after a tragic attack on a school in Minab that killed 170 students, 120 Democratic lawmakers formally asked the Pentagon whether an AI system, including the Maven Smart System, was used to identify the target. The Pentagon denied this, emphasizing that AI only assists in sifting information, with final decisions made by humans. Researchers warn that these systems are not sufficiently developed for warfare decisions.
The second problem is the ownership of these AI models by private tech giants like Google, Amazon, and Tesla. Their profit-driven motives clash with military ethics and accountability. Militaries are becoming dependent on corporate-controlled infrastructures, raising questions about who controls, profits from, and is accountable for these systems. The goals of these multinationals—maximizing revenue—often override ethical considerations, transparency, and error tracking.
Errors and Biases in AI Systems
Beyond decision-making, AI systems suffer from training database issues where input data differs from training data, leading to errors in novel situations. Automation bias, user bias, and interaction errors further compound the risks. As the world evolves, AI models matching inputs only to historical training data can produce flawed results.
The urgent need for stringent regulations on companies, militaries, and states is clear. Policymakers must address this impending threat to prevent warfare from falling into the hands of unaccountable corporate entities. The writer, Iman Shahbaz, is a law student at Lahore University of Management Sciences with interests in religion and history.
