The United States and Israel are deploying artificial intelligence tools to conduct warfare against Iran with unprecedented speed and precision, striking over 3,000 targets since last Saturday. AI is instrumental in gathering intelligence, selecting targets, planning bombing missions, and assessing battle damage, enabling commanders to manage supplies and select optimal weaponry at speeds previously unattainable.
Despite the clear benefits in managing complex military operations, the use of AI in warfare is not without its limitations and risks. U.S. military investigators suspect American forces may have been responsible for a strike that killed dozens of children at a girls' elementary school in Iran on the war's first day, highlighting the potential for severe consequences from poorly informed decisions, even with AI assistance. The quality of intelligence and analysis remains dependent on factors like imagery clarity and weather conditions.
The adoption of AI in the military is a key initiative for Defense Secretary Pete Hegseth, who advocates for an "AI-first" warfighting force. This push has led to public disputes, notably between Hegseth and Anthropic, a critical AI supplier to the Pentagon. Despite President Trump's order to cease using Anthropic's products and the Pentagon's subsequent contracting with rivals like OpenAI and Google, U.S. officials acknowledge the utility of Anthropic's AI agent, Claude, in the ongoing conflict. Anthropic has sued the Pentagon, asserting retaliation for its concerns about unchecked government use of its technology, including for autonomous armed drones and mass surveillance. Military leaders, such as Admiral Frank Bradley of US Special Operations Command, have also urged caution regarding AI's deployment in lethal capacities, emphasizing the need for safeguards to prevent unintended effects like civilian casualties.
