In a stunning revelation amid the ongoing U.S.-Israeli military campaign against Iran, The Wall Street Journal has reported that U.S. forces in the Middle East continued to rely on Anthropic’s Claude AI tool for operations — even hours after President Donald Trump ordered a full government ban on the company’s technology.
The strikes, part of Operation Epic Fury targeting Iranian leadership (including the confirmed death of Supreme Leader Ayatollah Ali Khamenei on February 28, 2026), involved commands worldwide, particularly U.S. Central Command (CENTCOM) in the region. According to people familiar with the matter cited by WSJ, Claude was actively in use during planning and execution phases of the attacks, despite the presidential directive issued just prior.
On February 27, 2026, Trump directed all federal agencies to immediately cease using Anthropic products, with a six-month phase-out for the Pentagon due to deep integration in classified systems.
Trump’s post stated: “We don’t need it, we don’t want it, and will not do business with them again!” The move escalated after Anthropic refused to lift restrictions by a Pentagon deadline, leading to the company being designated a potential national security risk.
Despite the order, operational inertia and embedded systems meant Claude remained active in military workflows during the critical hours leading into the February 28 strikes on Iranian targets, including Khamenei’s Tehran headquarters.