Anthropic has disabled access to its newly released Fable 5 and Mythos 5 artificial intelligence models after receiving an abrupt export‑control directive from the U.S. government. The order requires the company to block all foreign nationals from using the systems, including foreign‑national employees working inside the United States. To ensure full compliance, Anthropic shut down both models for every customer worldwide.
The directive arrived late Friday and cited national security authorities but did not include detailed evidence explaining the concern. According to Anthropic, officials indicated awareness of a narrow “jailbreak” technique that could prompt Fable 5 to scan a specific codebase for software vulnerabilities. The company said it reviewed the demonstration and found only a small number of previously known, minor flaws—issues it says other publicly available models can already identify without any bypass.
Anthropic emphasized that no universal jailbreak has been discovered and noted that Fable 5 underwent extensive red‑teaming with U.S. and U.K. partners before launch. The company publicly disagreed with the government’s interpretation, warning that recalling a model over a narrow, non‑universal jailbreak could set a precedent that halts frontier‑model deployment across the industry.
Access to all other Anthropic models remains unaffected. The company says it is working to restore availability but must comply with the directive while discussions with federal officials continue.
Sources
- Yahoo
- CNBC
- Android Authority
- POLITICO
- Bloomberg
- Anthropic