Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Probably—but the headline needs qualification. The Guardian reported that the U.S. military used Anthropic’s Claude in connection with the opening of its Iran operation, which began on February 28, 2026, one day after President Donald Trump ordered federal agencies to stop using Anthropic technology. That report does not establish that Claude independently selected a target or authorized a strike. The Defense Department was also given a six-month phaseout for existing use, so the announcement did not mean every deployed system shut off immediately.
What happened, and when
The overlap between the order and the Iran strikes is real, but it is not proof that the Pentagon simply ignored an immediate, technically complete shutdown. The key dates are:
![Claude AI for Beginners Bible: [5 in 1] The Ultimate Guide to Automate Your Work, Save Hours Every Week, and Use AI for Real-World Results](https://m.media-amazon.com/images/I/415+fSJacsL._SL160_.jpg)
- February 26, 2026: Anthropic CEO Dario Amodei said the company could not accept Pentagon demands for unrestricted military use of Claude.
- February 27: Trump directed federal agencies to cease using Anthropic technology. The Pentagon also moved to designate Anthropic a supply-chain risk. The administration described a six-month phaseout for Defense Department systems already dependent on the company’s technology.
- February 28: U.S. forces began Operation Epic Fury, including strikes in Iran.
- March 1: The Guardian reported that Claude was used in connection with the operation, citing people familiar with the matter rather than a publicly released Pentagon operational record.
The Guardian’s report supports saying Claude was reportedly still in use around the opening of the campaign. It does not provide a public audit establishing exactly which version was used, what task it performed, or how its output affected a particular decision. The Guardian’s report and Defense One’s account of the directive describe different parts of that picture.
What did Trump’s “ban” mean?
“Trump banned Claude” compresses several actions that should be kept separate. The president issued a public directive to federal agencies to stop using Anthropic technology. The Pentagon separately took procurement and supply-chain action against Anthropic. And the announced six-month transition acknowledged that Defense Department systems already depended on the technology.
#1 Best Overall
That is not the same as a single switch that instantly removed Claude from every government or contractor environment. Existing contracts, classified networks, cloud services and software integrations can make a technology difficult to replace quickly. Before the directive, Defense One reported that replacing Anthropic tools on classified networks could take months or longer. That reporting helps explain why a transition period mattered; it does not prove that every use was allowed, or that no alternatives existed.
The supply-chain designation also had implications beyond whether a federal office could directly open a Claude interface: it could affect contractors and suppliers working with the government. The directive, procurement restrictions and technical removal of already-integrated tools are related, but they are not interchangeable. A March court filing reproduced the administration’s position on the six-month phaseout; it was the government’s account in litigation, not a final judicial ruling on every use during the transition. The filing and GSA’s February 27 statement provide context for the federal response.
Rank #2
“Used Claude” can mean very different things
In a military setting, a model may be one component of a larger system rather than a stand-alone tool making operational decisions. Depending on the deployment, a model could help analyze or summarize intelligence, classify information, support planning or contribute to a contractor-integrated platform. Those functions are materially different from generating a target recommendation, choosing a weapon, or controlling a weapon.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Reports described Claude as present in classified military and intelligence environments through a wider ecosystem of cloud and defense-technology providers, including contractor-managed platforms. That does not mean Anthropic directly operated a strike system. A typical chain can involve the model developer, the cloud provider, a contractor integrating the software, military users and commanders responsible for decisions. Without details about the specific deployment, “Claude was used” does not identify which link in that chain it occupied.
Did Claude cause the Minab school strike?
The public evidence does not establish that it did. The Guardian reported that the Shajareh Tayyebeh primary school in Minab was struck on the first morning of Operation Epic Fury. Later reporting cautioned against the simplified claim that Claude itself generated the target package or caused the strike.
That reporting described the military’s Maven Smart System as the system used to produce targeting recommendations and recommend courses of action, including weapons and platforms. It does not publicly resolve whether Claude contributed to any part of that chain. In particular, the available record does not show that Claude independently identified the school, that an Anthropic model hallucinated it as a target, or that the model had authority to release weapons. The Guardian’s later analysis addresses the distinction between the wider targeting system and Claude.
To assess responsibility, it helps to separate the stages: intelligence collection; data processing and analysis; target-package preparation; human review; command authorization; weapon and platform selection; strike execution; and post-strike assessment. Public reporting leaves important parts of this chain unclear. An AI system’s presence somewhere in a workflow is not evidence that it made, approved or executed the final decision.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Why Anthropic objected—and what the Pentagon argued
Anthropic said it supported national-security work and had deployed Claude in classified government networks, but opposed two uses: mass domestic surveillance of Americans and fully autonomous weapons operating without meaningful human control. The company’s position was not that the military should never use AI; it argued that those applications should remain outside its government contracts. Amodei’s statement on the discussions sets out the company’s account.
The Pentagon’s stated position was that the Defense Department, rather than a private AI company, must control lawful military operations. It argued that negotiating model-specific restrictions could make a system unreliable in combat or other operational settings. The administration characterized Anthropic’s limits as an unacceptable attempt to dictate military policy; Anthropic characterized the demands as an effort to remove safeguards against mass surveillance and fully autonomous weapons. Those are the parties’ competing arguments, not a settled finding about the law or the particular Iran operation. AP’s report on Anthropic’s objections and its report on Trump’s directive outline the dispute.
The legal fight did not end with the announcement
Anthropic sued the administration, arguing that the government’s actions were retaliatory and unlawful. The government denied unlawful retaliation. On April 3, GSA said it was restoring Anthropic to the prior status quo after a federal judge paused implementation of the administration’s directive. That pause was an interim development, not a final judgment resolving the dispute. The legal status should therefore be described as contested rather than as a settled, permanent ban. GSA’s statement on the preliminary injunction describes the agency’s response.
What remains unknown
- Which Claude model or version was involved, and whether it was accessed directly or through a contractor-integrated platform.
- The precise task Claude performed during the Iran operation.
- Whether its output entered the targeting chain connected to the Minab school, and if so, how.
- Which personnel reviewed intelligence, recommendations and targeting decisions, and who authorized the strike.
- Whether the relevant use took place before the directive was operationalized, and how the phaseout applied to that specific system.
Until operational records or more specific evidence answer those questions, claims that Claude chose the school, caused civilian deaths or acted autonomously go beyond what the public reporting establishes.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
