WASHINGTON, United States — The United States government has cleared Anthropic to make its Claude Mythos 5 artificial intelligence model available again to a limited group of approved organisations.
The decision, made on Friday, June 27, 2026, partially eases restrictions imposed this month over national security concerns.
The decision allows selected companies and infrastructure providers to regain access to Mythos 5, which Anthropic describes as its most powerful cybersecurity model.
It does not extend to Fable 5, a less powerful model that remains covered by the earlier restrictions.
In a letter dated Friday, June 26, 2026, Howard Lutnick, the Commerce Secretary, said the company had worked with the government since the export block was issued to reduce concerns around the affected models.
Since the restrictions were imposed, “Anthropic has worked with the US government to address risks associated with the Covered Models,” Lutnick wrote.
“In light of progress in that work,” he added, “I have determined that appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model.”
Anthropic Says Access Will Be Restored
Anthropic confirmed that it had received notice from the government allowing a limited redeployment of Mythos 5.
“We received notice from the US government that Mythos 5, our strongest cybersecurity model, can be redeployed to a small group of cyber defenders and infrastructure providers,” Anthropic said in a statement.
“We are working to provision the approved set of providers and restore their access to Mythos 5 as quickly as possible. We are pleased to see this progress and continue to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again,” the statement added.
A person familiar with the talks told CNN that discussions between Anthropic and the government were expected to continue into the weekend, including over whether access to Fable could also be restored.
Earlier this month, Anthropic disabled customer access to both Mythos and Fable after the government ordered the company to suspend use by foreign nationals.
The restriction applied broadly enough to include some Anthropic employees.
AI Rules Remain Unsettled
The dispute has highlighted the absence of a stable regulatory framework for advanced artificial intelligence systems, even as the United States seeks to maintain its lead over competitors including China.
The Trump administration had previously taken a largely light-touch approach to artificial intelligence regulation.
But the action against Anthropic came after months of concern among experts that Mythos could present a security risk if used by hackers or other hostile actors.
The government has also asked OpenAI to restrict access to its newest model, according to the reporting notes.
Anthropic’s relationship with the federal government had already been strained. Earlier this year, the administration labelled the company a “supply chain risk” after a disagreement over military uses of its products.
Anthropic sued over that designation and secured at least one early ruling in the continuing case.
Semafor first reported the government’s decision to permit the limited release of Mythos.
