Anthropic Refuses Pentagon Demand to Drop AI Safeguards, $200 Million Contract at Risk

The dispute centers on the Pentagon’s demand that Anthropic allow its AI models to be used without restrictions that currently prevent the technology from supporting fully autonomous weapons or conducting mass domestic surveillance. The company has resisted those changes, even after being warned it could be labeled a “supply chain risk” and removed from Defense Department systems, per a report by Reuters.

Earlier in the day, Pentagon spokesperson Sean Parnell addressed the issue on X, stating that the department does not seek to deploy AI for widespread surveillance of Americans or to develop weapons that operate entirely without human oversight. “Here’s what we’re asking: Allow the Pentagon to use Anthropic’s model for all lawful purposes,” Parnell said. He added that the company had until 5:01 p.m. ET on Friday to comply. “Otherwise, we will terminate our partnership with Anthropic and deem them a supply chain risk.”

Anthropic Chief Executive Dario Amodei reiterated the company’s opposition to enabling its AI systems for mass domestic surveillance or fully autonomous weapons. In a statement released Thursday, Amodei said that “frontier AI systems are simply not reliable enough” to power such weapons. He emphasized that the company’s stance reflects a safety assessment rather than an accusation that the Pentagon intends to pursue those applications.

A source close to Anthropic, cited by Reuters, said the company believes advanced AI models remain too unpredictable for “life-or-death targeting,” warning that unexpected behavior in unfamiliar situations could result in “friendly fire, mission failure or unintended escalation.” The same source raised concerns about domestic surveillance, arguing that current laws do not limit the inferences AI systems can generate from vast datasets. As a result, AI could create broad population-level profiles that “no law explicitly prohibits but that clearly violate the spirit of constitutional protections,” the source said.

Read more: Anthropic’s Discovery of Autonomous AI Espionage Raises Legal and Regulatory Questions

Amodei indicated that while he hopes the Pentagon reconsiders its position, Anthropic is prepared for the possibility of losing the contract. The company “will work to enable a smooth transition to another provider,” he said, should the department move forward with termination.

According to Reuters, Amodei also said the Pentagon had threatened to remove Anthropic from its systems, designate it a supply chain risk and potentially invoke the Defense Production Act to compel removal of the safeguards. “Regardless, these threats do not change our position: we cannot in good conscience accede to their request,” he said.

The dispute has sparked sharp criticism from within the Defense Department. Undersecretary of Defense Emil Michael responded on X, writing, “It’s a shame that @DarioAmodei is a liar and has a God-complex. He wants nothing more than to try to personally control the US Military and is ok putting our nation’s safety at risk.” Michael added that the Pentagon would “ALWAYS adhere to the law but not bend to whims of any one for-profit tech company.”

Anthropic, which counts Google and Amazon among its backers, holds a Defense Department contract valued at up to $200 million. An Anthropic spokesperson said the company remains “ready to continue talks and committed to operational continuity for the Department and America’s warfighters.”

The standoff has drawn broader attention within the tech industry. More than 200 employees from Google and OpenAI signed an open letter supporting Anthropic’s position, according to Reuters. Google and OpenAI did not immediately respond to a Reuters request for comment.

Source: Reuters