1 hour ago
5
Anthropic CEO Dario Amodei said successful an exclusive interrogation with CBS News Friday that the Pentagon's determination to designate the AI institution a proviso concatenation risk was "retaliatory and punitive."
Hours earlier, Defense Secretary Pete Hegseth deemed the AI institution a "supply concatenation hazard to nationalist security," which restricts subject contractors from doing concern with Anthropic. The determination came aft Anthropic refused to give the subject unfettered entree to its AI model.
Amodei told CBS News elder concern and exertion analogous Jo Ling Kent that the actions were "unprecedented" and that "it was made precise wide successful immoderate of their statements, successful immoderate of their language, that this was retaliatory and punitive." Hei noted that Hegseth's designation was the archetypal ever issued for a U.S. company.
He said the institution sought to gully "red lines" successful the government's usage of its exertion due to the fact that "we judge that crossing those lines is contrary to American values, and we wanted to basal up for American values."
"Disagreeing with the authorities is the astir American happening successful the world," Amodei said. "And we are patriots. In everything we person done here, we person stood up for the values of this country."
The CEO said that if helium could talk to President Trump straight astir the dispute, helium would stress that Anthropic is made up of "patriotic Americans."
"Everything we person done has been for the involvement of this country, for the involvement of supporting U.S. nationalist security," Amodei said. "Our leaning guardant successful deploying our models with the subject was done due to the fact that we judge successful this country."
In July 2025, the Pentagon awarded Anthropic a $200 cardinal declaration to make AI capabilities that would beforehand U.S. nationalist security.
However, the Defense Department and the artificial quality institution person been successful negotiations implicit concerns from the AI startup that its model, Claude, could perchance beryllium utilized for surveillance connected Americans and the improvement and usage of autonomous weapons. Anthropic sought definite guardrails implicit the Pentagon's usage of Claude, which it said were rejected.
On Friday, Mr. Trump successful a societal media post ordered each national agencies to "immediately" halt their usage of Anthropic's technology. Mr. Trump said immoderate agencies, including the Defense Department, would person six months to signifier retired their usage of Anthropic.
Hegseth followed that up aboriginal Friday by declaring successful his ain societal media station that Anthropic would beryllium deemed a "supply concatenation hazard to nationalist security," and said that nary contractor who does concern with the Pentagon tin proceed to behaviour commercialized enactment with Anthropic.
The Pentagon had earlier this week fixed Anthropic a deadline of 5:01 p.m. to either scope a woody oregon suffer its authorities contracts.
Emil Michael, the Pentagon's main exertion officer, told CBS News connected Thursday that the subject had "made immoderate precise bully concessions" to Anthropic. He aboriginal added that "at immoderate level, you person to spot your subject to bash the close thing."
Watch the afloat interrogation with Anthropic CEO Dario Amodei Saturday greeting connected CBSNews.com oregon connected our YouTube channel.
In:
Anthropic CEO connected "red lines" for AI subject usage
Anthropic CEO connected "red lines" for AI subject use: "We wanted to basal up for American values"
(01:09)
English (US) ·