Anthropic’s Claude Fable 5 Returns With Stricter Usage Limits
▼ Summary
– Claude Fable 5 is available globally starting July 1, while Mythos 5 is available to select US organizations after government approval.
– US Secretary of Commerce Howard Lutnick stated the government worked with Anthropic for two weeks to analyze and approve Fable 5, ensuring alignment and strengthening US AI leadership.
– Anthropic addressed a vulnerability reported by Amazon with a new safety classifier that blocks 99% of malicious attempts, rerouting blocked requests to Opus 4.8.
– The classifier may flag benign requests during coding tasks, and Anthropic plans to reduce false positives.
– Fable 5 access includes up to 50% weekly usage limits through July 7, then switches to usage credit pricing, with access restored on AWS, Google Cloud, and Microsoft Foundry soon.
The United States government and Anthropic have confirmed that Claude Fable 5 is now available starting Wednesday, July 1. This advanced model is being rolled out to users worldwide, while a separate version, Mythos, will be accessible to a select group of organizations within the United States.
United States Secretary of Commerce Howard Lutnick shared the news on Twitter, stating: “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.” Unsurprisingly, his announcement drew complaints from those frustrated by the delay.
Anthropic Addresses Vulnerability With New Classifier
According to Anthropic’s official release, the vulnerability previously reported by Amazon has been resolved through a new classifier,an automated security layer designed to differentiate between harmless user prompts and malicious attempts to bypass guardrails. The company claims this updated system should catch 99% of such attempts.
When a request to Fable 5 is blocked, users will receive a notification, and the system will automatically reroute the query to Opus 4.8. Anthropic moved quickly to address the reported bypass, working closely with government officials to train an improved classifier that specifically targets the behavior described in the report.
Still, the company acknowledges that the safeguard is not perfect. Their announcement warns that “benign requests” may occasionally be flagged “during routine coding and debugging tasks.” Anthropic has promised to continue refining the classifier to reduce false positives over time.
Fable 5 Debuts With Stricter Usage Limits
Access to Fable 5 comes with weekly usage limits set at up to 50% through July 7. After that date, the model will shift to a usage credit pricing model.
Anthropic’s official statement laid out the access timeline clearly: “Fable 5 will be available starting tomorrow, Wednesday, July 1, to users globally on the Claude Platform, Claude.ai, Claude Code, and Claude Cowork. For Pro, Max, Team, and select Enterprise plans, Fable 5 will be included for up to 50% of weekly usage limits through July 7, after which it will be available via usage credits. We will re-enable access on AWS, Google Cloud, and Microsoft Foundry as quickly as possible. We have also restored access to Mythos 5 for a set of US organizations, following the US government’s approval on June 26. We continue to coordinate with the government to expand access to the broader set of domestic and international partners in the Glasswing program.”
(Source: Search Engine Journal)
