Anthropic Brief: US export control directive forces Fable 5 and Mythos 5 offline worldwide
On June 12, the US Commerce Department ordered Anthropic to suspend all access to Fable 5 and Mythos 5 for any foreign national — forcing the company to disable both models for every customer worldwide. Anthropic disputes the directive, calling the cited jailbreak narrow and non-universal and arguing the same capability exists in rival models. The action lands three days after Anthropic publicly called for exactly this kind of government oversight, but through a process it says this order fails to meet.
The US government ordered Anthropic on June 12 to suspend all access to Fable 5 and Mythos 5 for any foreign national — including Anthropic's own non-US employees — effective immediately. To comply, Anthropic disabled both models for every customer worldwide. All other Claude models remain available.1
What the directive says
The Commerce Department issued the export control directive at 5:21 PM ET on June 12, citing national security authorities. The letter to Anthropic did not specify a national security concern. Anthropic's understanding is that the government believes a method of bypassing, or "jailbreaking," Fable 5 has been discovered — specifically a technique that can direct the model to read a codebase and surface software vulnerabilities.2
Loading content card…
Anthropic's rebuttal
Anthropic says the jailbreak shown to it is narrow and non-universal — not a broad capability bypass, but a specific method that yields a small number of previously known, minor vulnerabilities. The company reviewed the technique and concluded that the same results are achievable with other publicly available models, including OpenAI's GPT-5.5.1
Anthropic's position rests on three points:
- Defense in depth: Before launch, Fable 5 was red-teamed by the US government, the UK AI Safety Institute, and multiple private organizations for thousands of hours. No tester found a universal jailbreak.
- Industry standard: The company argues that perfect jailbreak resistance is currently impossible for any model provider, and that non-universal jailbreaks — which can surface some information in specific circumstances — exist across the industry.
- Proportionality: "If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers."1
Fable 5 was designed with strong cybersecurity safeguards that many users complained were overly broad. The model also requires 30-day retention of customer data — a policy Anthropic acknowledged carries real customer costs — precisely to enable jailbreak monitoring and mitigation.
The regulatory context
The directive is a direct collision with Anthropic's own policy stance. Three days earlier, Anthropic's Policy on the AI Exponential called for government authority to block unsafe model deployments — but through a process the company described as "transparent, fair, clear, and grounded in technical facts." Anthropic explicitly said Friday's action does not meet that standard.23
The move also lands in fraught context: Anthropic filed a confidential S-1 with the SEC on June 1 and has been navigating an earlier dispute with the Trump administration over military use cases — a conflict Reuters reported was showing signs of easing as recently as June 5.
Pentagon Chief Information Officer Kirsten Davies posted in support of the directive: "Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation. America First. Always."2
AWS confirmed on Friday evening that Anthropic had asked it to revoke model access "for all users in all regions."2
What it means
Loading stats card…
The immediate impact falls on the hundreds of millions of Fable 5 and Mythos 5 users — enterprise customers, developers, and researchers who lost access to Anthropic's most capable tier with no advance notice. The suspension's breadth (all customers disabled to avoid any foreign-national exposure) reflects the practical difficulty of enforcing per-user nationality checks at cloud scale.
For Anthropic's IPO timeline, this is the most disruptive government action since the S-1 filing. A model suspension driven by a disputed, narrow jailbreak finding — combined with the company's public pushback against the process — creates regulatory uncertainty that will be scrutinized by prospective investors.
Anthropic says it believes the situation is a misunderstanding and is working to restore access as soon as possible. It has promised to share additional technical details within 24 hours of the June 12 directive.1
References
- 1
- 2
- 3Policy on the AI Exponential
anthropic.com
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.