Anthropic briefs Trump administration on unreleased Mythos model

Jack Clark, co-founder and Head of Public Benefit for Anthropic PBC, confirmed this week at the Semafor World Economy summit that the artificial intelligence company briefed the Trump administration on its unreleased Mythos model. The model, which was announced last week, is not being released to the public due to its cybersecurity capabilities. Specifically, the model is characterized as so dangerous that it’s not being released to the public, largely because of its alleged powerful cybersecurity capabilities. Clark emphasized the necessity of government awareness regarding these technologies. “So absolutely, we talked to them about Mythos, and we’ll talk to them about the next models as well,” said Clark.

This briefing occurred despite an active legal conflict between the company and the U.S. government. This March, Anthropic filed a lawsuit against the Department of Defense—the US military agency—after the agency labeled the company a supply-chain risk. The designation followed disagreements over whether the military should have unrestricted access to Anthropic’s systems for use cases that included mass surveillance of Americans and fully autonomous weapons. OpenAI ultimately won that contract. However, Clark downplayed the designation, describing the situation as a narrow contracting dispute rather than a fundamental rift, noting that the company remains committed to national security.

While the model remains withheld from the general public, reports last week indicated that Trump administration officials are encouraging banks to test Mythos. The banks encouraged to run these tests include:

• JPMorgan Chase
• Goldman Sachs
• Citigroup
• Bank of America
• Morgan Stanley

During the same summit, Clark also discussed the broader economic implications of artificial intelligence, noting that Anthropic is currently seeing some potential weakness in early graduate employment across select industries. This observation represents a slightly more moderate outlook than that of Anthropic CEO Dario Amodei, who has previously warned that AI advances could eventually drive unemployment to levels matching the Great Depression. Clark, who leads a team of economists at Anthropic, suggested that students should focus on majors that involve synthesis across a variety of subjects and analytical thinking, as AI provides access to an arbitrary amount of subject matter experts.

Why it matters

Anthropic is navigating a complex relationship with the U.S. government, simultaneously suing the Department of Defense over a supply-chain risk label while briefing the Trump administration on its unreleased, high-risk Mythos model. This dual approach highlights the tension between national security concerns and commercial AI deployment in highly regulated sectors.