Anthropic CEO Dario Amodei was reportedly set to meet Friday with White House Chief of Staff Susie Wiles, in the wake of a bitter feud in which the Pentagon blacklisted the AI giant and the exec lashed out at the Trump administration.
The West Wing meeting was expected as leaders in the US and beyond have been grappling with Anthropic’s Mythos, a new artificial-intelligence model that the company itself says could create widespread online disruption due to its ability to breach cybersecurity defenses.
“It would be grossly irresponsible for the US government to deprive itself of the technological leaps that the new model presents,” a source close to the talks told Axios. “It would be a gift to China.”
The White House confab comes as Anthropic has been in discussions to provide government agencies with advance access to the system, according to the Wall Street Journal. The administration has convened meetings involving Vice President JD Vance, Treasury Secretary Scott Bessent and Defense Secretary Pete Hegseth, along with senior execs including Amodei, to coordinate preparations for Mythos.
Instead of a public release, Anthropic has unveiled “Project Glasswing,” a plan to share Mythos with a handpicked group of about 40 companies including Amazon, Apple, CrowdStrike, Google, JPMorgan Chase and Nvidia. They are receiving early access to the model so they can use it to find and fix security flaws.
Anthropic and the White House have feuded for months over safety restrictions.
After the company refused to remove safeguards blocking its AI models from being used to power autonomous weapons or mass surveillance of Americans, the Pentagon last month labeled the company a supply-chain risk — barring it from working with the Department of War and prohibiting contractors from using Anthropic’s AI models in their work with the Pentagon. The White House directed all federal agencies to drop Anthropic’s tech, too.
Meanwhile, Amodei lashed out at the president in a leaked internal email, accusing the Department of War of targeting Anthropic for not giving “dictator-style praise to Trump.” The exec later apologized, as his company fought the blacklisting in court.
Still, the feds have been prepping a version of Mythos for key agencies, according to Bloomberg. Recent days have seen staff from two large agencies contact Anthropic about integrating Mythos into their cyber-defense projects, Politico reported.
Anthropic has been working with lobbyists with White House ties as it tries to ease the tensions, according to the Journal.
Following the blacklisting, the company hired mega-lobbyist Brian Ballard, the publication reported. He raised over $50 million for Trump’s 2024 presidential run. The company works with Carlos Trujillo, a veteran of the first Trump White House, too.
The Pentagon reached agreements with Anthropic rivals OpenAI and xAI to use their models in classified settings. But security analysts say it could take months before those systems are integrated into operations at the level Anthropic’s technology has achieved.
