US lifts Anthropic's Claude Mythos 5 ban for 100+ firms

Key Takeaways

• US Commerce Department reversed its June 12 order, allowing Anthropic to release Claude Mythos 5 to over 100 trusted partners including Fortune 500 companies
• Export licenses no longer required for approved companies and their foreign national employees to access the AI model
• Fable 5, another restricted Anthropic model, remains under review with no clear timeline for release

The US Commerce Department has reversed course on Anthropic, allowing the AI company to release its Claude Mythos 5 model to more than 100 "trusted partners" just two weeks after ordering a complete suspension. Commerce Secretary Howard Lutnick signed the authorization on Friday, according to a letter seen by Reuters.

The original June 12 order blocked access to Anthropic's most advanced models over fears that military intelligence users in China, Russia, and other adversarial nations could exploit them. Anthropic responded by disabling both Mythos 5 and another model called Fable 5 for all users, not just foreign ones.

Who gets access to Mythos 5 now?

The approved list includes over 100 companies and institutions, many of them Fortune 500 firms, according to a source familiar with the directive who spoke on condition of anonymity. Export licenses will no longer be required for these trusted partners or their foreign national employees.

Companies not on the approved list still face restrictions. The letter did not publish the full roster of trusted partners, leaving unclear which specific firms made the cut.

"In just two weeks, we have worked diligently to ensure America remains the global leader in AI while safeguarding our security," a Commerce Department spokesperson said.

What happened to Fable 5?

Lutnick's letter made no mention of Fable 5, the other model Anthropic disabled after the June 12 order. A source told Reuters that the government is moving toward allowing its release as well, but provided no timeline.

This leaves Anthropic in a strange position: one flagship model cleared for commercial use, another still in regulatory limbo. For enterprise customers evaluating AI providers, that uncertainty matters.

Why did the government act so quickly?

Two weeks is fast for export control reversals. The speed suggests Anthropic addressed the Commerce Department's specific concerns about access controls and monitoring, though neither party disclosed what safeguards were implemented.

"Since the issuance of my June 12 letter, Anthropic has worked with the U.S. government to address risks associated with the Covered Models. These efforts have yielded significant progress," Lutnick wrote.

The Commerce Department has increasingly treated frontier AI models as dual-use technology, similar to advanced semiconductors or encryption tools. This framework subjects them to export controls designed to prevent adversaries from acquiring militarily useful capabilities.

Anthropic's rocky relationship with Washington

The Mythos suspension is not Anthropic's first conflict with the federal government this year. The company refused to allow the US military to use its AI models for domestic surveillance and fully autonomous weapons systems. According to the Economic Times report, the government retaliated by placing Anthropic on a national security blacklist.

That tension sits awkwardly alongside Anthropic's position as a leading American AI company. Backed by $4 billion from Amazon and $2 billion from Google, Anthropic competes directly with OpenAI and is reportedly preparing for an IPO. Being labeled a security risk while pursuing a public offering creates obvious complications.

Neither Anthropic nor the White House commented on the new directive.

What this means for AI procurement

Enterprise buyers now face a new variable when selecting AI providers: export control risk. A model available today could be suspended tomorrow if the Commerce Department decides it poses national security concerns. The Mythos episode shows these suspensions can be reversed quickly, but the two-week blackout disrupted customers regardless.

Companies building production systems on frontier models may want to assess whether their AI vendors have the government relationships and compliance infrastructure to navigate these situations. OpenAI, Google, and Anthropic all face similar regulatory exposure as their models grow more capable.

Frequently Asked Questions

What is Claude Mythos 5?

Mythos 5 is described as one of Anthropic's most advanced AI models. It was suspended from distribution on June 12 over concerns about potential use by adversarial nations, then cleared for release to trusted partners on June 28.

Which companies can access Anthropic Mythos 5?

Over 100 companies and institutions, including many Fortune 500 firms, are now authorized to use the model. The Commerce Department has not published the full list of approved partners.

Is Fable 5 available to use?

No. The Commerce Department's letter addressed only Mythos 5. Fable 5 remains restricted, though sources indicate the government is working toward its release.

Do foreign employees need special licenses to use Mythos 5?

Not if they work for approved companies. The new directive exempts foreign national employees of trusted partners from export license requirements.

Source: Tech-Economic Times / ET