Anthropic’s co-founder confirmed that the company informed the Trump administration about the Mythos
Jack Clark, an Anthropic co-founder and the company’s Head of Public Benefit, confirmed that Anthropic has briefed the Trump administration regarding its latest AI model, Mythos.
Unveiled last week, the model is being withheld from public release due to significant safety concerns, primarily stemming from its allegedly powerful cybersecurity capabilities.
Speaking at this week’s Semafor World Economy summit, Anthropic co-founder Jack Clark addressed the apparent contradiction of the company actively engaging with the U.S. government while simultaneously pursuing legal action against them.
In March, Anthropic initiated legal action against the Trump administration’s Department of Defense (DOD) following the agency’s designation of the company as a “supply-chain risk.” The dispute centered on conflicting views regarding military access to Anthropic’s AI technology, particularly concerns over its potential use in mass surveillance and fully autonomous weaponry. Ultimately, the Pentagon chose to award the contract to OpenAI instead.
During the conference, Clark minimized the significance of the administration’s “supply-chain risk” designation, characterizing the situation as merely a “narrow contracting dispute.” He emphasized that the company remains deeply committed to national security and does not want this disagreement to overshadow that objective.
“Our stance is that the government must remain informed about these developments,” Clark stated. He emphasized the necessity of forging new collaborative frameworks between the government and the private sector, particularly as companies build transformative AI technologies that intersect with national security and societal interests. “Consequently, we have briefed them on Mythos, and we intend to maintain that dialogue regarding our future models as well.”
This confirmation follows reports from last week indicating that Trump administration officials have been encouraging major financial institutions—specifically JPMorgan Chase, Goldman Sachs, Citigroup, Bank of America, and Morgan Stanley—to test the Mythos model. During the same interview, Clark also discussed broader societal implications of AI, touching upon its potential impact on unemployment rates and the future of higher education.
While Anthropic CEO Dario Amodei has previously cautioned that rapid AI advancements could trigger unemployment rates comparable to the Great Depression, Jack Clark offers a slightly more nuanced perspective. Clark explained that Amodei’s projections are based on the premise that AI will achieve significant power much faster than most anticipate, thereby driving his more conservative economic forecasts.
Jack Clark, who leads the economics team at Anthropic, stated that the company has observed only “some potential weakness in early graduate employment” within select industries thus far. Nevertheless, he emphasized that Anthropic remains proactively prepared to address any significant shifts in the labor market should they occur.
When pressed for specific advice on which college majors students should prioritize or avoid in an AI-driven landscape, Clark refrained from naming particular fields. Instead, he offered a broad recommendation, suggesting that the most valuable academic paths are those that cultivate analytical thinking and the ability to synthesize information across a wide range of disciplines.
“That’s because AI effectively grants you access to a nearly limitless array of subject matter experts across various domains,” Clark explained. “However, the truly critical skill is knowing how to ask the right questions and possessing the intuition to identify what happens when you combine insights from disparate disciplines.”
