U.S. Orders Anthropic to Suspend Fable 5 and Mythos 5 Access for Foreign Nationals
U.S. Orders Anthropic to Suspend Fable 5 and Mythos 5 Access for Foreign Nationals
https://thehackernews.com/2026/06/us-orders-anthropic-to-suspend-fable-5.html
Publish Date: 2026-06-13 01:42:00
Source Domain: thehackernews.com
Summary:
Anthropic faced a sudden directive from the U.S. government to disable its most advanced AI models, Fable 5 and Mythos 5, for foreign nationals due to alleged national security concerns. The company argues that the government’s claim of a potential “narrow, non-universal jailbreak” method used to bypass Fable 5’s safeguards is simplistic and similar to vulnerabilities found in other AI models. Anthropic believes there’s a “misunderstanding” and claims that its models have effective and superior safeguards compared to previous models. While the export control directive affected the advanced models, the less strictly regulated Mythos 5 remains accessible for specific cybersecurity-related users. The company emphasizes the impracticality of a “perfect” jailbreak resistance and believes that a narrow discovery shouldn’t warrant such a significant response.
Key Points:
- Anthropic disabled its advanced AI models, Fable 5 and Mythos 5, due to U.S. government orders citing national security concerns.
- The company disputes the government’s claim of a potential bypass method, asserting that other models possess similar vulnerabilities.
- Mythos 5, still operational for vetted cyber defenders, utilizes the same underlying model but with heightened cybersecurity safeguards.
- Anthropic maintains that while it supports addressing security concerns, the directive based on a “narrow potential jailbreak” might not justify disabling such widely deployed commercial models.
- The company argues for a transparent and evidence-based process when dealing with concerns over AI deployments.
