The United States government has allowed approximately 100 American companies and government agencies to access Anthropic’s most powerful artificial intelligence model, according to the news portal Semafor.
According to unidentified sources, the Department of Commerce announced in a letter to Anthropic that it was lifting restrictions on the use of Mythos 5 for a select list of companies and federal departments, after its utilization was suspended nearly two weeks ago over national security concerns.
The letter does not explain what will happen to the Fable 5 model, a less advanced version of Mythos that had been made publicly available on the Claude platform and also saw its access suspended.
“I have determined that the necessary safeguards are in place to allow certain trusted partners to access Claude Mythos 5,” US Commerce Secretary Howard Lutnick stated in the letter.
Read more about this topic: Japan demands explanations from US amid restrictions on Anthropic AI models (with video)
According to the US government, Anthropic demonstrated “significant progress” following several days of negotiations prompted by fears that the company’s most advanced models could be used in cyberattacks.
This news emerged on Friday, the same day OpenAI unveiled its new GPT-5.6 Sol, Terra, and Luna models, which also feature limited access restricted to partners approved by the US government.
In a statement, OpenAI expressed opposition to government approval “becoming the long-term standard,” as it “deprives users, developers, enterprises, cybersecurity professionals, and global partners of the best tools.”
On June 13, Anthropic suspended public access to its most advanced AI models to comply with an export control directive, which requires the company to block access to the service by foreign nationals for national security reasons.
Read more about this topic: Fujitsu expands AI strategy through collaborations with OpenAI and Anthropic
According to Anthropic, to comply with the order to block access outside the United States or by foreign nationals within the country, it chose to cut off all access to Fable 5 and Mythos 5 for all customers.
In a statement, Anthropic noted that the US government did not provide specific details regarding why the new artificial intelligence services pose a national security issue, and apologized to customers for the disruption.
“As we have stated publicly, we believe the government should have the capacity to block unsafe developments as part of a transparent, fair, clear, and technically fact-based process. This action is not aligned with those principles,” the company stated.
The suspension occurred just days after Anthropic made the Fable 5 model available to subscribers of its Claude service. Fable 5 had been presented as the most advanced model across various capability tests and was considered significantly superior to the Opus models.
