The Trump administration has spent nearly two months fighting with AI company Anthropic. It’s dubbed the company a “RADICAL LEFT, WOKE COMPANY” full of “Leftwing nut jobs” and a menace to national security. But some of the ice may reportedly be melting between the two, thanks to Anthropic’s buzzy new cybersecurity-focused model: Claude Mythos Preview.
Anthropic’s relationship with the Pentagon soured quickly in late February after the company refused to budge on two red lines: using its technology for domestic mass surveillance or lethal fully autonomous weapons with no human in the loop. Anthropic’s tech has in the past been used heavily by the DoD and, it was the first company to have its models cleared to operate on classified military networks. The stalemate led to public insults on social media, Anthropic being categorized as a “supply chain risk,” the company filing a lawsuit fighting that designation, and a temporary injunction halting its ban.
Anthropic has recently attempted to get back in the US government’s good graces, at least in some capacity, with Mythos Preview. And judging from reports that Anthropic CEO Dario Amodei attended a meeting at the White House on Friday, it may be working. (Anthropic did not provide a comment.)
Mythos Preview was announced with major fanfare about its capabilities — including the ability to find security issues in virtually every large web browser and operating system. Anthropic says the model is its most powerful yet, and it’s currently only available for private access. It’s being marketed as a way to flag high-stakes vulnerabilities in some of the most-used internet infrastructure we have, so that companies like Apple, Nvidia, and JPMorgan Chase — which have already signed on to use it — can plug them up before bad actors can exploit them. The release of Mythos Preview has already reportedly sparked emergency meetings between US bank leaders and Federal Reserve Chairman Jerome Powell.
The Trump administration, too, seems to be taking notice. In a release about Mythos Preview, Anthropic wrote that it had already been in “ongoing discussions with US government officials about Claude Mythos Preview and its offensive and defensive cyber capabilities.” Earlier this month, when The Verge asked for details, Dianne Penn, a head of product management at Anthropic, confirmed that the company had “briefed senior officials in the US government about Mythos and what it can do,” and that the company is still “committed to working closely with all different levels of government.” The company declined to specify who, exactly, had been briefed.
Anthropic also reportedly recently hired Ballard Partners, a lobbying firm linked to Trump, which has inspired more reports that a deal between Anthropic and the White House may be in the works.
On Friday, Axios reported that Amodei was scheduled for a meeting with White House chief of staff Susie Wiles later that day. Describing the reasons for the meeting, a source familiar with the negotiations said “it would be grossly irresponsible for the U.S. government to deprive itself of the technological leaps that the new model presents” and that “it would be a gift to China.” The outlet also reported that “some parts of the U.S. intelligence community, plus the Cybersecurity and Infrastructure Security Agency (CISA, part of Homeland Security)” are testing Mythos Preview, and that other departments and agencies are interested.
If Amodei did attend the meeting and it opens up conversations about further integrating Anthropic’s Claude into government usage across agencies, it’s possible that the DoD could shift its views on Claude accordingly as well. It would be an anticlimactic end to a bitter fight over national security — but hardly the first time the administration has suddenly reversed course.
