Anthropic is reinstating worldwide access to its premier AI models, Claude Fable 5 and Claude Mythos 5, after the U.S. Department of Commerce lifted export restrictions imposed due to security concerns earlier this year.
- U.S. export ban on Fable 5 and Mythos 5 lifted starting July 1
- Access restored globally via Anthropic's Claude platform and select partners
- New safeguards implemented to address previous security vulnerabilities
What happened
Anthropic announced on July 1 that it would restore access to its AI models Claude Fable 5 and Claude Mythos 5 globally. This follows the U.S. government’s decision to lift export controls that had previously blocked foreign nationals from using these models due to their advanced capabilities and an identified security vulnerability.
The ban originated after researchers at Amazon found a flaw that allowed users to circumvent Fable 5's safeguards. Although Mythos 5 was never widely available, access was also restricted. Since then, Anthropic has implemented a new classifier and updated safeguards to prevent dangerous usage and reduce risk.
Why it matters
Fable 5 and Mythos 5 represent some of the most advanced AI capabilities available today, with Mythos 5 particularly adept at identifying security vulnerabilities. Restoring access enables a wider range of users, from individuals to enterprise customers, to leverage these powerful tools for development and security analysis.
The episode highlights ongoing tension between innovation in AI technology and national security concerns. Anthropic’s coordinated efforts with government regulators to address vulnerabilities and ensure safe deployment serve as a model for balancing advancement with responsible usage.
What to watch next
Anthropic plans to continue expanding access through its Glasswing program, working with both domestic and international partners to further deploy Mythos 5. The company also aims to make Fable 5 available on major cloud providers like AWS, Google Cloud, and Microsoft Foundry soon to reach broader enterprise markets.
Future updates are expected to refine the new safeguard measures, which presently result in some false positives during routine coding and debugging. Monitoring how Anthropic balances robust security controls with user experience will be critical in assessing the ongoing safe use of these advanced AI systems.