The US Commerce Department has authorized Anthropic to resume limited deployment of its Claude Mythos 5 model to approximately 100 companies and federal agencies, partially reversing the export control directive that suspended all access two weeks ago. The decision, communicated in a letter from Commerce Secretary Howard Lutnick to Anthropic cofounder Tom Brown, does not extend to the company’s Fable 5 model, which remains restricted.
The Commerce Department Letter
“I have determined that appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model,” Lutnick wrote in the letter, according to CNBC, which viewed the document. The letter cited “significant progress” in Anthropic’s work with the government to address risks associated with its most advanced models.
The roughly 100 approved organizations include Fortune 500 companies and operators of critical infrastructure, according to Reuters, citing a source familiar with the directive who declined to be identified.
Anthropic confirmed the development in a statement: “Today, the government notified us that Mythos 5, our strongest cybersecurity model, can be redeployed to a set of US organisations that operate and defend critical infrastructure. We’re restoring access for these organisations quickly, and we’re continuing to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again.”
Timeline of the Standoff
Anthropic disabled access to both Mythos 5 and Fable 5 on June 12 after receiving an export control directive from the Trump administration citing “national security authorities.” The order required Anthropic to suspend all access “by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.”
The suspension came days after Anthropic announced the models, which the company described as state-of-the-art across multiple benchmarks. Two weeks of negotiations followed, with cofounder Tom Brown reportedly replacing CEO Dario Amodei as Anthropic’s lead negotiator with the White House. NCT covered the stalled state of those negotiations on June 26, when the standoff appeared at risk of extending indefinitely.
Friday’s letter breaks the deadlock, but only partially. The selective “trusted partner” model means most Anthropic customers and developers still cannot access Mythos 5. Fable 5 access remains fully suspended.
Criticism of the Vetting Process
The government’s approach of selecting which companies receive access has drawn pushback from both industry and civil liberties groups.
“No one knows how these companies are picked and why everyone else is excluded,” John Coleman, legislative counsel for the Foundation for Individual Rights and Expression, told Reuters. “This is putting too much power in the hands of the government. There’s little transparency and it raises questions about the rule of law.”
OpenAI CEO Sam Altman also weighed in on X, acknowledging that extensive safety testing “is not a bad idea” but adding, “I just don’t like the idea of the government picking the customers.” This came on the same day OpenAI announced it was delaying a full public launch of GPT-5.6 at the government’s request, limiting access to its own set of vetted partners.
A Parallel Governance Framework Emerges
The simultaneous restrictions on both Anthropic and OpenAI’s latest models point to a new regulatory pattern: the US government is moving toward customer-by-customer access approval for frontier AI systems rather than blanket bans or unrestricted availability. Both companies now operate under similar “trusted partner” frameworks where federal officials vet who can use their most advanced models.
The practical question is what this means for speed. Enterprise customers on the approved list can now deploy Mythos 5 for cybersecurity and critical infrastructure applications. Everyone else waits for the government to expand the list, with no published timeline for broader access or for the Fable 5 restrictions to lift.