Anthropic made ‘mistake’ in Pentagon talks: FCC Chair
FCC Chairman Brendan Carr testifies before the House Appropriations Subcommittee on Financial Services and General Government at the Rayburn House Office Building in Washington, DC, on May 21, 2025.
John McDonnell | Getty Images
Anthropic “made a mistake” in its dealings with the Department of Defense, Federal Communications Commission Chairman Brendan Carr told CNBC on Tuesday, after the US government blacklisted the artificial intelligence firm.
Anthropic was in tense negotiations with the Pentagon over the terms of its contract. The startup sought assurances that its technology would not be used for fully autonomous weapons or domestic mass surveillance of Americans. The Department of Defense requested that Anthropic allow the military to use the models in all legal use cases.
Talks stalled last week, and Anthropic CEO Dario Amodei said the company “cannot in conscience allow” its models to be used under such conditions.
“According to me [Anthropic] The FCC’s Carr told CNBC he probably made a mistake. “Clearly there are rules of the road that apply to every technology for which the War Department contracts.”
President Donald Trump later ordered all US government agencies to “immediately cease” using Anthropic’s technology. Defense Secretary Pete Hegseth turned up the pressure on Anthropic, labeling it a “Supply Chain Risk to National Security.” This designation means that any contractor working with the Pentagon will not be able to do business with Anthropic.
Asked by CNBC if the door was still open for Anthropic to work with the U.S. government, the FCC’s Carr said the company “should try to course-correct as much as it can.”
“They were given a lot of landing pads… a lot of opportunities to find a great landing spot and they chose not to do that and that’s a mistake on their part,” Carr added.
Antropik was not immediately available for comment when contacted by CNBC.
Anthropic on Friday in question He was “saddened” by the blacklisting, saying it would “be both legally unsound and set a dangerous precedent for any American company negotiating with the government.”
“We have sought to reach an agreement in good faith with the War Department, making clear that we support all lawful uses of artificial intelligence for national security, except for mass domestic surveillance of Americans and fully autonomous weapons,” Anthropic said.
Just hours after Anthropic was blacklisted, OpenAI CEO Sam Altman said his company had reached an agreement with the Department of Defense regarding the use of artificial intelligence models. On Monday, Altman said OpenAI “shouldn’t rush” its deal with the Department of Defense, adding that the deal “seems opportunistic and sloppy.”
OpenAI outlined revised terms of the agreement, including language clarifying that “the AI system will not be used intentionally for domestic surveillance of U.S. citizens and citizens.”
– CNBC’s Ashley Capoot and Dylan Butts contributed to this report.
