U.S. Central Command (CENTCOM) said it was using artificial intelligence tools to quickly process large amounts of data during strikes on Iran. The idea was to automate the initial selection of signals and leave the verification and final decisions to humans.
U.S. Central Command (CENTCOM) said it was using artificial intelligence tools to quickly process large amounts of data during strikes on Iran. The idea was to automate the initial selection of signals and leave the verification and final decisions to humans.
According to Bloomberg, the US has struck more than 2,000 targets since the campaign began, including about 1,000 in the first 24 hours. CENTCOM commander Admiral Brad Cooper called it almost «twice the scale» of the 2003 US campaign in Iraq.
Command spokesman Capt. Timothy Hawkins says AI helps analysts narrow focus, generate «points of interest» and organize information within systems. He emphasized that the tools do not select targets or make kill decisions — that goes through a separate command and legal process consistent with U.S. policy and military doctrine.
The sources also said that among the systems used in the operation is the Maven Smart System, a digital mission management platform from Palantir that collects data from more than 150 sources. They said that it also employs Anthropic’s Claude, which has become a central element in accelerating the analysis.
In parallel, the Pentagon’s conflict with Anthropic over the terms of use of the models continues. Defense Secretary Pete Hegseth called the company a risk to the supply chain and gave contractors six months to stop working with it; President Donald Trump has instructed federal agencies to curtail work with Anthropic. Negotiations between the parties have resumed.
The use of AI in warfare is fueling debates about automation bias, where humans overly trust systems’ advice. CENTCOM is also investigating reports of possible civilian casualties after a school strike that witnesses said killed more than 160 people.
Previously, dev.ua wrote about how Anthropic removed safeguards in its own Claude security rules after pressure from the Pentagon.
