Anthropic CEO Dario Amode said Thursday that the company “cannot in good conscience” allow the Defense Department to use its models without limits in all legitimate use cases, adding that the agency’s threats would not change its position.
The artificial intelligence startup has engaged in tense talks with the Pentagon in recent weeks, and those discussions are still ongoing, it said Thursday. Defense Secretary Pete Hegseth has threatened to label Anthropic a “supply chain risk” or invoke the Defense Production Act to force the company to comply with his demands.
Anthropic wants assurances that its models won’t be used for fully autonomous weapons or mass domestic surveillance of Americans, but the DOD wants to use the models without those restrictions.
“It is the department’s prerogative to select the contractor that best aligns with its vision,” Amode wrote in a statement. “But given the substantial value that anthropogenic technology provides to our armed forces, we hope they will reconsider.”
Hegseth met with Amode at the Pentagon on Tuesday, and he gave Anthropic until Friday evening to agree to his agency’s demands. DOD sent Anthropic its “last and final proposal” Wednesday night, according to a senior Pentagon official.
Chief Pentagon spokesman Sean Parnell said Thursday that the DOD has “no interest” in using anthropic models for fully autonomous weapons or conducting mass surveillance of Americans, which he noted is illegal. He emphasized that the agency wants the company to agree to allow its samples to be used for “all legitimate purposes”.
“This is a simple, common-sense request that prevents Anthropic from jeopardizing critical military operations and putting our warfighters at risk,” Parnell wrote Thursday in a post on X.
Anthropic signed a $200 million contract with the DOD in July and is the first lab to integrate its models into mission workflows on classified networks.
The startup’s competitors OpenAI, Google and xAI were awarded contract awards of up to $200 million from the DOD last year. Those companies have agreed to allow the DOD to use their models for all legal purposes in the military’s unclassified systems, though xAI agreed this week to allow its models to be used in classified settings.
“Our strong priority is to continue serving the department and our warfighters — with our two requested safeguards,” Amode said. “If the department chooses Offboard Anthropic, we will work to enable a smooth transition to another supplier, avoiding any disruption to ongoing military planning, operations or other critical operations.”
Watch: Anthropic’s AI Safety Contradiction
(tags to translate) Breaking News: Technology
