NewsBite analysis · AI, Security
Anthropic
Redeploying Claude Fable 5
Anthropic has lifted export controls on its Claude Fable 5 and Mythos 5 models after addressing cybersecurity concerns that led to their suspension. Users can now access Fable 5 globally, with usage limits on select plans. The company has implemented additional safeguards to prevent the exploitation of vulnerabilities while working closely with the US government on compliance and safety measures.
MonitorConfidence: MediumShort term horizon
Published 1 Jul 2026, 04:21 UTC14 min read
Visit original article ↗Lead image — Anthropic
Summary
TL;DR
- Export controls on Claude Fable 5 and Mythos 5 were lifted after government concerns.
- Access to Fable 5 will be restricted initially for certain pricing plans.
- New safety classifiers are introduced to detect and prevent harmful cybersecurity tasks.
- The collaboration with the US government aims to ensure compliance and enhance model security.
- A shared industry framework is being developed to address AI model jailbreaks.
Interesting facts
- The lifting of export controls allows Claude Fable 5 to become more accessible internationally.
- Colleagues within Anthropic were assigned to bolster the safety team prior to the model’s launch.