Anthropic CEO meets White House amid Pentagon blacklist

Washington D.C. — Anthropic CEO Dario Amodei met White House Chief of Staff Susie Wiles on Friday, April 17, 2026, as the AI company remains in a legal fight with the Trump administration over the Pentagon’s treatment of its Claude model.

 

The meeting took place while Anthropic is challenging the Pentagon’s decision to blacklist Claude as a “supply chain risk.” Officials have framed the issue around national security, while the situation also underscores the government’s interest in advanced AI tools that could support federal missions.

 

According to the account of the dispute, Claude had previously been the only AI model operating inside the Pentagon’s classified network. That relationship later deteriorated after the Trump administration moved to cut ties when Anthropic declined to provide the military with unrestricted use of Claude for “all lawful purposes,” including autonomous weapons and mass surveillance.

 

The breakdown led Anthropic to sue the administration. A federal judge in California temporarily blocked the blacklisting last month, and the government has said it intends to appeal that ruling. The court action leaves the status of the Pentagon’s restrictions unsettled, with the next steps dependent on the appeals process.

 

Even as the Claude dispute continues, the White House is also exploring access to a newer Anthropic model called Mythos. The model is described as being designed to identify cybersecurity threats, placing it in a category of tools that can be relevant to both civilian agencies and national security functions.

 

The Office of Management and Budget is preparing to provide agencies with access to Mythos, and the White House is reportedly in discussions to obtain access as well. That parallel track suggests the administration is weighing competing priorities: limiting perceived risks tied to one model while still seeking to use other Anthropic systems for government needs.

 

For markets and policy watchers, the episode illustrates how procurement decisions, security designations, and court challenges can shape the commercial outlook for leading AI developers. It also highlights how government demand for AI—particularly in cybersecurity—can persist even amid litigation over separate use cases.

 

Key uncertainties remain. The government’s planned appeal could change the legal posture around the Pentagon’s blacklist, while discussions over Mythos access indicate that interagency decisions on deployment and permissions are still evolving.