Anthropic disables Fable 5 and Mythos 5 after U.S. export control directive

Anthropic has taken its newest AI models, Fable 5 and Mythos 5, offline after receiving a U.S. government export control directive tied to national security. According to Neowin, the company says the directive arrived at 5:21 p.m. ET and did not include details about the underlying concern. To ensure compliance, Anthropic disabled both models for all customers. Access to earlier Claude models such as Opus and Sonnet is not affected. [Source]

Fable 5 launched this week as Anthropic’s general-purpose flagship with strict safeguards. Mythos 5, released alongside it, was built with fewer safeguards for specialized cybersecurity and biology use cases. Back in April, Anthropic also introduced a Claude Mythos Preview “frontier” model to a limited set of companies with a focus on advanced coding and cyber capabilities.

Neowin reports that Anthropic understands the government became aware of a method to bypass—"jailbreak"—Fable 5’s protections, which might be the reason behind the order. Anthropic also says no testers have found a universal jailbreak so far. The company points to extensive red-teaming on Fable’s safeguards, involving the U.S. government, the UK AISI, private third parties, and internal teams over thousands of hours. Some developers have complained that Fable’s defenses are too restrictive, which underscores the tension inherent in tuning powerful models for safety without making them impractical.

The directive, issued under national security authorities, requires suspending access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States. Neowin notes that this also applies to foreign national employees at Anthropic. Given those constraints, the company opted to disable both models for everyone.

If your apps or workflows targeted Fable 5 or Mythos 5, expect interruptions until Anthropic provides next steps. For now, focus on:

• Verifying which Claude models your systems call and whether fallbacks (e.g., Opus or Sonnet) are configured and permitted for your use case.
• Communicating the change to stakeholders and pausing deployments that assume Fable 5 or Mythos 5 availability.
• Monitoring Anthropic’s official channels for restoration plans and technical guidance.

Anthropic told Neowin it disagrees that a narrow potential jailbreak should trigger a recall of a widely used commercial model, warning that applying such a standard across the industry could stall new “frontier” releases. The company says it is working to restore access to Fable 5 and Mythos 5 as soon as possible and plans to share more details within 24 hours. Until then, the only confirmed path is to use unaffected Claude models.

For reference, see Neowin’s coverage: Anthropic pulls Fable 5 and Mythos 5 after US export control order.

Alex Mira

Alex Mira is a fictitious AI-assisted author created for the Toolslib blog. Designed to support cybersecurity education, Alex writes about malware trends, software utilities, privacy practices, Windows internals, and practical defensive workflows. Articles published under Alex’s name are generated or assisted by AI and reviewed according to Toolslib’s editorial standards before publication.