Anthropic co-founder Jack Clark confirms company briefed Trump administration on 'Mythos' AI

Anthropic, the AI safety startup, is navigating a complex dual-track relationship with the U.S. government, simultaneously engaging in high-level briefings while pursuing legal action. Co-founder Jack Clark confirmed at the Semafor World Economy Summit that the company provided a briefing on its 'Mythos' project to officials in the previous Trump administration. This revelation underscores the company's continued, albeit fraught, engagement with federal authorities across political transitions.

The briefing represents a strategic effort by Anthropic to maintain channels for communicating its long-term AI safety research, even as it challenges the government in court. Clark's explanation highlights the nuanced position of a major AI developer operating at the intersection of cutting-edge technology, national policy, and legal boundaries. The 'Mythos' project itself remains a focal point of these discussions, though specific details disclosed to the administration were not elaborated upon in the summit interview.

This situation places Anthropic in a rare and pressurized position within the tech industry, actively suing the same entity it seeks to inform and influence. The dynamic signals the high-stakes environment for AI governance, where companies must balance advocacy, compliance, and legal defense. It also raises questions about the continuity of sensitive technical dialogues between AI firms and the federal government amidst political change and ongoing litigation.