Image Credits:Chris Ratcliffe/Bloomberg / Getty Images7:56 AM PDT · April 18, 2026
Despite precocious being designated a supply-chain consequence by nan Pentagon, Anthropic is still talking to high-level members of nan Trump administration.
There were earlier signs of a thawing narration — aliases a consciousness that not each portion of nan management wanted to trim disconnected Anthropic — pinch reports saying that Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell were encouraging nan heads of awesome banks to trial retired Anthropic’s caller Mythos model.
Anthropic co-founder Jack Clark seemed to corroborate this, claiming that nan ongoing conflict complete nan supply-chain consequence nickname is simply a “narrow contracting dispute” that would not interfere pinch nan company’s willingness to little nan authorities astir its latest models.
Then connected Friday, Axios reported that Bessent and White House Chief of Staff Susie Wiles had met pinch Anthropic CEO Dario Amodei. In a statement, nan White House described this arsenic an “introductory meeting” that was “productive and constructive.”
“We discussed opportunities for collaboration, arsenic good arsenic shared approaches and protocols to reside nan challenges associated pinch scaling this technology,” nan White House said.
Similarly, Anthropic issued a statement confirming that Amodei had met pinch “senior management officials for a productive chat connected really Anthropic and nan U.S. authorities tin activity together connected cardinal shared priorities specified arsenic cybersecurity, America’s lead successful nan AI race, and AI safety.”
The institution added that it’s “looking guardant to continuing these discussions.”
Techcrunch event
San Francisco, CA | October 13-15, 2026
The conflict betwixt Anthropic and nan Pentagon seemingly began aft grounded negotiations complete nan military’s usage of Anthropic’s models; nan AI institution sought to support safeguards astir nan usage of its exertion for afloat autonomous weapons and wide home surveillance. (OpenAI quickly announced a subject woody of its own, starring to immoderate consumer backlash.)
The Pentagon subsequently declared Anthropic a supply-chain risk — a explanation that’s mostly reserved for overseas adversaries and could severely limit nan usage of Anthropic’s models by nan government. The institution is challenging that nickname successful court.
But it sounds for illustration nan remainder of nan Trump management doesn’t stock nan Pentagon’s hostility, pinch an management root telling Axios that “every agency” isolated from nan Department of Defense wants to usage nan company’s technology.
Anthony Ha is TechCrunch’s play editor. Previously, he worked arsenic a tech newsman astatine Adweek, a elder editor astatine VentureBeat, a section authorities newsman astatine nan Hollister Free Lance, and vice president of contented astatine a VC firm. He lives successful New York City.
You tin interaction aliases verify outreach from Anthony by emailing anthony.ha@techcrunch.com.
16 hours ago
English (US) ·
Indonesian (ID) ·