The Iran War Highlights the Dangers of AI Involvement in Military Decision-Making

According to a report by Reference News on March 26, the Spanish website "Confidential" reported on March 21 that artificial intelligence (AI) played a very important role in the military strikes by the United States and Israel against Iran.

“60 seconds, that’s all it took.” A former Israeli Mossad agent described the attack on February 28, which resulted in the death of Iran's Supreme Leader Khamenei on the first day of the US-Israel operation against Iran. The use of AI systems significantly increased the speed and scale of warfare. However, this pursuit of speed poses serious risks to both civilians and military personnel.

Modern military operations rely on and generate vast amounts of intelligence information, including intercepted phone calls and text messages, large-scale internet surveillance, satellite imagery, and drone videos. We can view this information as data, but the problem lies in the sheer volume of data.

General Brad Cooper, commander of the US Central Command, recently confirmed the use of AI tools in the conflict with Iran, stating: “These systems help us sift through massive amounts of data in seconds, allowing us to filter out noise and make smarter decisions faster than the enemy.”

During World War II, the selection cycle for aerial targets could take weeks or even months. However, in the following decades, the US military focused on what is known as "compressed kill chains," which aim to shorten the time from target identification to the use of force. Shortly after the 9/11 attacks, armed "Predator" drones emerged, reflecting significant advancements by the US military in this area.

These drones are equipped with powerful cameras that can capture high-resolution video and transmit it back to the United States via satellite within seconds, allowing operators to track moving targets.

The use of AI tools may partly stem from a belief that human cognition cannot compete with the processing speed offered by AI systems. The US Department of Defense noted: “Military AI will be a race for the foreseeable future, so speed is key.”

Although the specific use of AI by the US and other militaries is kept confidential, information regarding civilian risks has been made public, highlighting the dangers posed to civilians by AI systems.

According to Israeli intelligence sources, during operations in Gaza, the Israeli AI system was set to accept casualties of up to 100 civilians in a single strike against suspected Hamas militants.

The risk is that as personnel become increasingly passive in reviewing and approving AI operations, AI may lower the threshold or cost of waging war. Accountability becomes increasingly difficult to achieve.

The integration of AI into the military strike chain intertwines with other concerning developments. Infrastructure designed to avoid civilian casualties in warfare has been dismantled, and lawyers responsible for advising on target selection actions and ensuring that military operations comply with international law have been marginalized or dismissed.

US Secretary of Defense Lloyd Austin has made it clear that military targets in Iran are “maximum lethality, not a meaningless legality,” and that the focus is on “pursuing violent effects rather than political correctness.” With this attitude, the US military will prioritize speed over caution, making civilian casualties inevitable and accountability increasingly elusive. (Translated by Han Chao)