Anthropic said the US government has lifted export controls on its Claude Fable 5 and Mythos 5 artificial intelligence models.
The company plans to begin restoring access to both models from Wednesday.
Anthropic Restores Access to Fable 5 and Mythos 5
Anthropic confirmed the policy change in a blog post published on Tuesday evening.
The announcement followed an earlier report from Politico indicating that Washington was preparing to remove the restrictions.
Users will regain access to Fable 5 and Mythos 5 through the Claude Platform, Claude.ai, Claude Code, and Claude Cowork.
Cloud Access Will Also Return
Anthropic said it is also working to restore access through major cloud providers.
These include Amazon Web Services, Google Cloud, and Microsoft Foundry.
The company said it aims to complete the process as quickly as possible.
White House Order Previously Restricted the Models
Anthropic abruptly suspended public access to Fable 5 and Mythos 5 earlier in June.
The decision followed a White House directive that restricted foreign nationals from using the models.
According to Anthropic, the export controls were introduced after Amazon researchers discovered a possible method for bypassing some of Fable 5’s safety protections.
Anthropic Strengthens AI Safety Measures
Over the following two weeks, Anthropic worked closely with the US government and technology partners, including Amazon.
The company said the collaboration focused on improving the models’ safety systems and reducing potential misuse.
Last week, Washington allowed Anthropic to provide Mythos to a limited number of trusted US organizations.
That decision was viewed as an early sign that the government could reverse the broader restrictions.
Cybersecurity Risks Delayed Public Release
Anthropic had previously delayed the launch of Fable and Mythos because of concerns about their potential use in developing serious cybersecurity exploits.
The company consulted with several large US organizations and financial institutions before making the models publicly available in early June.
Anthropic said its latest assessment found that Fable 5 did not pose the cybersecurity risks that had initially concerned officials.
Anthropic Defends Fable 5 Safeguards
The AI company said Fable 5 was released with its strongest model safeguards to date.
Anthropic maintained that the improved protections were sufficient to address the concerns raised by government officials and research partners.
The lifting of the export controls will now allow Anthropic to expand access to the models across its own products and major cloud platforms.
