Anthropic Rejects Pentagon’s New Terms, Stands Firm on Lethal AI Weapons and Surveillance

Anthropic’s refusal triggered an immediate Pentagon warning that the $200 million contract would be terminated and the firm labeled a “supply‑chain risk,” according to a statement from the Department of Defense cited by The Guardian. The deadline was set for Friday, and CEO Dario Amodei replied less than 24 hours before it expired, saying the threats “do not change our position” (The New York Times).

Amodei reiterated that Anthropic cannot, in good conscience, strip safety checks from Claude or grant unrestricted access for lethal autonomous weapons and mass‑surveillance, a stance echoed in his own company statement (Anthropic). He emphasized the company’s commitment to defending democracies while maintaining ethical safeguards.

Pentagon officials, as reported by Reuters, have escalated the dispute, threatening to deem Anthropic a supply‑chain risk if it does not comply. The move could jeopardize the startup’s eligibility for future defense contracts and affect its standing with other federal agencies.

TechCrunch noted that the standoff may force the Pentagon to seek alternative AI vendors, potentially reshaping the defense AI market. Industry analysts warn that a prolonged impasse could slow the deployment of advanced AI tools across U.S. military platforms.

Despite the pressure, Anthropic’s leadership remains firm. The company points to its pioneering deployments on classified networks and at national laboratories as evidence of its commitment to national security without compromising core safety principles (Anthropic).