According to Anthropic, Fable 5 became available again on July 1 across Claude.ai, Claude Code, Claude Cowork, and the Claude API. Cloud integrations via AWS, Google Cloud, and Microsoft Foundry are also being restored.

The redeployment follows nearly three weeks of disruption after US authorities imposed emergency export controls on both Fable 5 and the more advanced Mythos 5 model on June 12.

Export Controls Triggered by Cybersecurity Concerns

The restrictions were introduced after government officials reviewed a report from Amazon researchers who discovered a way to bypass some of Fable 5’s safeguards and force the model to identify software vulnerabilities.

In one case, the model reportedly generated exploit code tied to a known vulnerability, raising concerns about offensive cybersecurity misuse.

Anthropic said its internal review found that several competing frontier models — including OpenAI GPT-5.5 and older Claude models — could perform similar tasks, suggesting the issue was not unique to Fable 5.

New Safeguards Added Before Relaunch

Before bringing the model back online, Anthropic says it worked directly with US authorities and industry partners to strengthen its security systems.

The company deployed updated safety classifiers specifically designed to block the exploit pathway identified in the original report. If a user triggers a blocked request, the system will now automatically reroute the prompt to Claude Opus 4.8 instead.

Anthropic described the updated system as part of a broader “defense in depth” strategy for preventing misuse of frontier AI.

Mythos 5 Access Still Limited

While Fable 5 has returned for global users, access to Claude Mythos 5 remains restricted.

Anthropic confirmed that Mythos 5 has only been reactivated for a select group of US-based organizations approved by the government, mainly through its Project Glasswing cybersecurity initiative.

The company said it continues working with regulators to expand access to trusted international partners.

AI Regulation Debate Continues to Intensify

The temporary suspension and redeployment of Fable 5 marks one of the clearest examples yet of direct government intervention in frontier AI deployment.

The incident highlights growing tensions between rapid AI innovation and national security oversight, especially as advanced models become increasingly capable in areas like vulnerability discovery, autonomous coding, and cyber offense.

For Anthropic, the return of Fable 5 restores one of its most important flagship models — but under tighter safeguards and closer government scrutiny than ever before.

Source: https://www.anthropic.com/news/redeploying-fable-5