Shadows in the Strike: US Military Reportedly Used Claude AI in Iran Operations Despite Trump Ban

In a stunning revelation of the deep integration of artificial intelligence in modern warfare, reports from The Wall Street Journal and Axios indicate that the U.S. military relied on Anthropic’s Claude AI during yesterday's massive joint U.S.-Israeli bombardment of Iran—just hours after President Trump had ordered an "immediate" federal ban on the technology.

The operation, which reportedly resulted in the elimination of Iran's Supreme Leader, Ayatollah Ali Khamenei, has highlighted a growing rift between the White House's political directives and the tactical realities of the "Department of War."

1. The Friday Ban vs. The Saturday Strike
The timeline of events reveals a chaotic 24-hour window where policy clashed with active combat operations:

Friday, Feb 27 (Morning): President Trump issues an executive order via Truth Social, branding Anthropic a "Radical Left AI company" and ordering all federal agencies to cease using its tools immediately.

Friday, Feb 27 (Evening): Defense Secretary Pete Hegseth designates Anthropic as a "supply chain risk," a label usually reserved for foreign adversaries like Huawei.
+1

Saturday, Feb 28 (Early Hours): U.S. Central Command (CENTCOM) launches a coordinated strike on Tehran. Despite the ban, sources claim Claude was still active within the command's intelligence and targeting systems during the mission.
+1

2. How Claude Was Used in the Tehran Mission
According to military insiders, the AI wasn't just a "chatbot" but a core component of the mission's orchestration layer:

Target Identification: Claude was used to synthesize vast amounts of satellite imagery and signal intelligence to pinpoint high-value targets, including the Supreme Leader's office.

Battlefield Simulation: Before the first jet took off, the military reportedly ran thousands of simulations through Claude to predict Iranian air defense responses and calculate "collateral damage" estimates.

Intelligence Assessment: The AI processed real-time data feeds during the strike to provide commanders with instant assessments of "strike effectiveness" (BDI).

3. The "Deep Integration" Problem
The Pentagon’s continued use of the tool despite a direct presidential order underscores why a six-month phase-out period was eventually clarified by the administration.

Embedded Workflows: Claude is currently woven into the classified networks of the CIA, NSA, and CENTCOM through partnerships with Palantir and Amazon Web Services.

The "Maduro" Precedent: Officials noted that Claude was also the primary intelligence engine used in the January 2026 capture of Venezuelan President Nicolás Maduro, proving its operational value in high-stakes "decapitation" missions.

The OpenAI Pivot: Even as the ban was announced, OpenAI CEO Sam Altman confirmed his company has already reached a deal with the Pentagon to deploy its models on classified networks to replace Anthropic, though experts warn a "seamless transition" could take months.

4. The Political Standoff: Safeguards vs. Sovereignty
The root of the ban lies in a fundamental disagreement over "Red Lines."

Anthropic’s Refusal: CEO Dario Amodei refused a Pentagon ultimatum to remove safety guardrails that prevent Claude from being used for mass domestic surveillance or fully autonomous "lethal" decisions.

The Administration’s Stance: Secretary Hegseth argued that "America’s warfighters will never be held hostage by the ideological whims of Big Tech," demanding "unrestricted access" for any lawful military purpose.

The "Woke AI" Label: President Trump has consistently characterized Anthropic's ethical constraints as a national security threat, claiming their "selfishness" puts American lives at risk.

Strategic Note: "The fact that CENTCOM used Claude for the Iran strikes after the ban shows that the military currently views the technology as 'mission critical.' You can’t just flip a switch on the brain of your targeting system in the middle of a war." — Defense Tech Analyst

5. Current Status of the Transition
Mourning in Tehran: Iran has declared a 40-day mourning period and promised "crushing retaliation" for the strike.

Anthropic’s Legal Challenge: Anthropic has vowed to fight its "supply chain risk" designation in court, calling the move "legally unsound" and a "dangerous precedent."

The 6-Month Countdown: The "Department of War" has until August 2026 to fully strip Anthropic technology from its stack and replace it with "more patriotic" AI services.