Anthropic's Claude Used in US Military Operations Against Iran Despite Ethical Dispute
Anthropic disputed with the US Department of Defense over safeguards for its Claude AI system in military operations, specifically opposing its use for mass surveillance and autonomous weapons. Despite the dispute, the US military reportedly continued using Claude in operations against Iran, with military officials confirming AI tools help analyze data for targeting decisions while humans make final choices.
What happened
Anthropic disputed with the US Department of Defense over the use of its Claude AI system in military operations. The conflict centered on Anthropic's insistence on safeguards that would prevent the military from using Claude for mass surveillance of US citizens or for fully autonomous weapons. The Pentagon demanded the ability to use Claude for 'all legal purposes' and stated that Anthropic's concerns were not relevant.
Despite the dispute, the US military continued using Claude in operations against Iran. Brad Cooper, head of US Central Command, confirmed in a video statement that the military was using advanced AI tools to analyze large quantities of data in seconds to help leaders make faster decisions, while stating that humans always make final decisions on targeting.
Reported impact
Experts from Oxford University warned that the episode 'reveals long-standing governance gaps in the integration of AI in military operations, gaps that predate this administration and will survive the current controversy.' Concerns have grown about whether reliance on AI for decision-making could cause errors in military operations.
Index analysis
Logan Graham, leader of Anthropic's red team analyzing worst-case scenarios including cyberattacks and biosecurity threats, highlighted a governance vacuum around AI integration in military operations. The dispute raised questions about how far decisions about irreversible and lethal actions are being delegated to machines and who decides how AI is used.
Governance controls
Anthropic sought to establish explicit safeguards preventing Claude's use for mass surveillance of US citizens and fully autonomous weapons systems. The Pentagon countered that US law already prohibits such surveillance and internal military policies restrict its conduct.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.