Fable 5, Mythos 5 restored globally; safety classifier hardened against reported bypass
Claude Fable 5 and Mythos 5 have been redeployed globally after export controls were lifted. The models were temporarily restricted on June 12 following a government directive, but access has now been restored as of June 30, 2026.
Availability:
- Fable 5 is available starting July 1 to users globally on Claude Platform, Claude.ai, Claude Code, and Claude Cowork
- For Pro, Max, Team, and select Enterprise plans, Fable 5 is included for up to 50% of weekly usage limits through July 7, then available via usage credits
- Access on AWS, Google Cloud, and Microsoft Foundry is being re-enabled
- Mythos 5 access has been restored for a set of US organizations following government approval
Safeguard Updates: Anthropicworked with the government to train an improved safety classifier targeting the behavior described in a reported bypass. The new classifier blocks the specific technique in over 99% of cases. Researchers from the US Department of Commerce's Center for AI Standards and Innovation (CAISI) have tested both prior and new safeguards and agree they are extraordinarily strong.
Industry Framework: Anthropicis partnering with Amazon, Microsoft, Google, and other Glasswing partners to develop a consensus framework for assessing the severity of AI jailbreaks. The framework scores jailbreaks on four criteria: capability gain, breadth of capability gain, ease of weaponization, and discoverability.
Government Collaboration: Anthropicis strengthening collaboration with the US government on:
- Pre-release government access and evaluation for models advancing the capability frontier
- Rapid information sharing on safeguards and jailbreaks
- Dedicated resources for joint research on AI security
- Working toward a common industry security and evaluation standard
Fetched July 9, 2026
