Talks Between Trump Administration and Anthropic End Without Resolution on AI Export Controls

Negotiations between Trump administration officials and Anthropic concluded Monday without lifting export controls on the company’s two most powerful AI models, Wired reported, leaving the path forward unclear.

Commerce Secretary Howard Lutnick sent Anthropic CEO Dario Amodei a letter on June 13 imposing export controls on Fable 5 and Mythos 5, banning any foreign national — including Anthropic’s own foreign national employees — from accessing the models anywhere in the world. The letter cited national security concerns but provided no specific rationale. Anthropic disabled both models for all users that evening to comply.

Read: Anthropic Suspends Fable 5 Access After US Government Export Control Order 

Silver47 Exploration Inc. — sponsored Sponsored · Silver47 Exploration Inc.

An unnamed company told the administration it could jailbreak Fable 5 to access Mythos-level capabilities. Officials tried to get Anthropic to pause the release — Anthropic says it was never told why — and sent the export control letter when those efforts failed.

Anthropic reviewed the technique and concluded it involves asking the model to read a codebase and identify software flaws — a capability also present in other publicly available models, including OpenAI‘s GPT-5.5, and one used routinely by cybersecurity defenders. 

“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,” the company said in a statement.

Monday’s talks involved the Commerce Department’s CAISI and ONCD units. Lutnick dialled in from the G7 summit in France. National Cyber Director Sean Cairncross did not personally attend. Commerce has expressed willingness to restore Fable 5 for consumer use if Anthropic resolves the jailbreak concerns — but the two sides remain apart on whether those concerns are valid.

Anthropic released Fable 5 on June 9, describing it as “Mythos-class” — a new capability tier the company said exceeded any model it had previously made publicly available.



Information for this story was found via the sources and companies mentioned. The author has no securities or affiliations related to the organizations discussed. Not a recommendation to buy or sell. Always do additional research and consult a professional before purchasing a security. The author holds no licenses.

Video Articles

8 Mining Stocks Our Viewers Asked Us to Review | Sept 14th-18th

Canada Is Finally Fast-Tracking New Mines | John Passalacqua – First Phosphate

Gold & Silver Just Added $2 Billion to This Project | Aya Gold – Bourmadine PEA

Recommended

Cardiol Expands MAVERIC Trial to Up to 150 Patients After Reaching Enrollment Target Early

Japan Gold Lands $50 Million in Committed Capital Through Solidcore Alliance

Related News

Book Authors Win $1.5B Settlement Against Anthropic On Copyright Suit

Anthropic’s $1.5 billion copyright settlement has received final court approval as a remedy for claims...

Tuesday, July 21, 2026, 11:49:52 AM

Anthropic Accuses Alibaba of Largest Known AI Extraction Campaign, Urges Congress to Act

Operators linked to Alibaba's Qwen AI lab ran nearly 29 million unauthorized exchanges with Claude...
Thursday, June 25, 2026, 01:53:42 AM

Anthropic Shares Surge in Demand as OpenAI Struggles on Secondary Market

Anthropic has emerged as the hot ticket in the AI investment space, with secondary market...

Wednesday, April 1, 2026, 12:08:11 PM

US Sanctions Canada-Based Sandvine For Enabling ‘Mass Web-Monitoring And Censorship’ In Egypt

The US Department of Commerce has added Canada-based Sandvine Inc., a computer networking company, to...

Tuesday, February 27, 2024, 07:45:02 AM

Cooking the Books? Lutnick Suggests ‘Messing With The GDP’

The US administration will redefine how the government calculates gross domestic product, a move that...

Monday, March 3, 2025, 02:04:00 PM