Claude Fable 5 was pulled offline after a U.S. directive, raising a harder question about AI safety rules.
Anthropic’s Claude Fable 5 was pulled offline just days after launch after the U.S. government directed the company to suspend access to Fable 5 and Mythos 5 for foreign nationals. In a June 12 statement, Anthropic said the order forced it to disable both models for all customers because it could not reliably comply while keeping access open for others. Access to other Claude models was not affected, according to the company.
The dispute came three days after Anthropic launched Claude Fable 5 and Claude Mythos 5 on June 9. Anthropic described Claude Fable 5 as a “Mythos-class” model made available for general use, while Mythos 5 was reserved for selected cyberdefense and infrastructure partners. The company said both models were priced at $10 per million input tokens and $50 per million output tokens.
The core issue is a claimed jailbreak. Anthropic said it understood the government’s concern to involve a method of bypassing Fable 5 safeguards and using the model to identify software vulnerabilities. The company said it had received only verbal evidence of a “narrow, non-universal jailbreak” and argued that similar capability is already available from other public AI models.
Government officials appear to see the risk differently. Semafor reported that the White House action was partly tied to concerns that a China-linked group may have accessed Mythos, citing a person familiar with the matter. Semafor also reported that Amazon had raised concerns about a jailbreak, while Anthropic said the White House had not raised Chinese access in its conversations with the company.
The episode creates a difficult public position for Anthropic. The company has called for stronger government oversight of powerful AI systems, and The AI Decode recently covered how Anthropic’s AI regulation plan pushes for more testing before major model releases. Yet Anthropic said this specific order was not transparent, fair or grounded in clear technical evidence.
The commercial impact may extend beyond developers who lost access. On the same day as the government order, Anthropic announced a TCS partnership that will bring Claude to 50,000 TCS employees across 56 countries and to clients in regulated sectors such as banking, healthcare and public services. That makes reliability and access rules more than a technical issue. They are now part of enterprise risk planning.
Claude Fable 5 is now a test case for AI regulation. The open question is whether governments can block risky model access quickly without turning every frontier AI release into a legal and geopolitical dispute.
