Anthropic CEO Dario Amodei met White House chief of staff Susie Wiles and Treasury Secretary Scott Bessent Friday, a visit described by officials as productive. The talks come after the Pentagon branded the company a supply chain risk over a disputed $200 million contract. U.S. leaders now seek to reopen the door to Anthropic’s model, Mythos, cut off from federal work.

Mythos, unveiled April 7, can spot software vulnerabilities and is the only AI model the Defense Department had on classified computers before the dispute. Skeptics feared the tool could weaponize autonomous systems, while Anthropic demanded safeguards that its models would not command lethal weapons or surveil Americans. The Pentagon’s “supply chain risk” label, first used against foreign firms, triggered a lawsuit.

If a compromise materializes, it will likely exclude the Pentagon, officials said, allowing Anthropic to re‑enter federal contracts while keeping the agency’s concerns separate. The meeting signals Washington’s willingness to balance innovation against security, a stance that could reshape how defense agencies procure AI. For investors, a deal would restore a $200 million opportunity and validate Anthropic’s market position in 2026.