Anthropic Sends Senior Technical Team To Washington After Export Controls Force Shutdown Of Top Models

Anthropic dispatched a delegation of senior technical staff to Washington over the weekend to press U.S. officials for a resolution after an executive action from the Trump administration forced the company to suspend access to its two most capable models, Fable 5 and Mythos 5.

What Happened

On Friday the administration issued an export-control order restricting foreign governments, companies and individuals from accessing Anthropic’s Fable 5 and Mythos 5 models. Because the restriction also swept in foreign-born Anthropic employees, the company concluded the only practical way to comply was to suspend access for all users. Other Anthropic models were not affected.

Who Went To Washington

Anthropic sent senior security researcher Nicholas Carlini, Logan Graham, head of the team that evaluates models for risk, and Dave Orr, the company’s head of safeguards. They met with government security researchers and officials to seek a de-escalation and a path forward. Representing the administration were Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross. Anthropic’s delegation also included co-founder and chief compute officer Tom Brown and public policy lead Sarah Heck.

Why The Controls Were Imposed

The new restrictions were prompted by research from Amazon scientists who reported that safeguards on Fable 5 could be partially bypassed — a finding that alarmed White House officials. Amazon CEO Andy Jassy brought the research to the attention of U.S. officials, including Treasury Secretary Scott Bessent, and President Trump personally authorized the export restriction, appointing Lutnick to coordinate the administration’s response.

Anthropic’s Response

Anthropic disputed the severity of the report. The company said it reviewed what it believes to be the underlying research and concluded the level of capability demonstrated is already available from other publicly deployed models, including OpenAI’s GPT-5.5, and is routinely used by cybersecurity defenders. In a statement Anthropic said:

“We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people.”

Reactions And Wider Impact

On Sunday a letter circulated among cybersecurity researchers urged the administration to reverse course, warning that the controls had “taken the best models away from defenders, created market uncertainty, and risked America’s AI leadership without any real risk to justify it.”

The shutdown comes as Anthropic filed a confidential IPO prospectus with the SEC earlier this month, disclosing a reported revenue run rate of $47 billion and a valuation of $965 billion. The company has previously experienced friction with U.S. authorities: the Department of Defense once designated Anthropic a supply-chain risk after talks collapsed, and Anthropic has an active lawsuit challenging that designation.

What’s Next

Both sides signaled interest in resolving the dispute, but it remains unclear what a negotiated settlement might look like or how long the suspension will last. The meetings in Washington will likely focus on technical mitigations, export-control scope, and how to allow defenders and trusted partners to access high-capability models without increasing national-security risks.