Fable 5 Stages a Comeback Tour After 18-Day Ban, Mythos 5 Opens the Encores 🌍

The U.S. Department of Commerce lifted export controls on Anthropic's Claude Fable 5 and Mythos 5 on June 30, ending an 18-day restriction that had taken both models offline since June 12 and clearing the way for global access to resume on July 1. The reversal closes roughly three weeks of negotiations between Anthropic and the White House, and the company confirmed it directly: "We've received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5. We'll begin restoring access tomorrow, and will share an update soon." Anthropic posted on June 30. Fable 5 launched on June 9 alongside the more restricted Mythos 5, which is built on the same underlying architecture but ships with additional safeguards for cybersecurity workloads and is limited to vetted Project Glasswing partners.

The controls were imposed after Amazon researchers reported a method of bypassing Fable 5's guardrails, prompting the model to identify software vulnerabilities and, in one instance, demonstrate an exploit. Because the directive applied to all foreign nationals worldwide and Anthropic said it had no real-time way to verify user nationality, the company disabled both models across Claude.ai, the API, AWS Bedrock, and partner platforms. U.S. Secretary of Commerce Howard Lutnick wrote on X that, "Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America's leadership in AI." White House Chief of Staff Susie Wiles added that the government's priority remains to "get the best [AI] tech deployed as quickly and safely as possible." Researchers at the Commerce Department's Center for AI Standards and Innovation reviewed the new and prior safeguards and called them extraordinarily strong.

Anthropic's internal testing found that Claude Opus 4.8, GPT-5.5, and Kimi K2.7 could identify the same vulnerabilities Fable 5 flagged in the Amazon report and reproduce the single exploit demonstration. The company is redeploying Fable 5 with a new set of classifiers designed to block the reported bypass in more than 99% of cases, with any flagged request automatically rerouted to Opus 4.8. Anthropic acknowledged the tighter filter will also flag more routine coding and debugging requests and said it will continue tuning to reduce false positives. Mythos 5 had already begun returning on June 26 to institutions cleared by the government and is scheduled to return globally on July 2.

Anthropic is also expanding collaboration with Amazon, Microsoft, Google and other Project Glasswing partners on a draft consensus framework for assessing the severity of AI jailbreaks, with planned pre-release model access, information sharing on misuse, and dedicated resources for joint research. Access details for Fable 5 vary by plan: Pro, Max, Team, and select Enterprise subscribers get Fable 5 included for up to half their weekly usage limits through July 7, after which it shifts to usage credits, with availability on AWS, Google Cloud, and Microsoft Foundry to follow "as quickly as possible." While the models sat offline, OpenAI's GPT-5.5-Cyber reportedly topped Mythos on the CyberG benchmark, and a separate AI researcher claimed to have jailbroken Fable 5 within 48 hours of its June 9 launch, sharing screenshots of bypassed guardrails.