AI Watch: US blocks global access to Anthropic’s Fable 5, Mythos 5 AI models amid security concerns

The U.S. government has blocked global access to Anthropic’s Fable 5 and Mythos 5 AI models, citing security concerns. Officials claim a jailbreak finding suggests Mythos 5’s cybersecurity capabilities could be exploited through Fable 5, prompting export restrictions. Anthropic has responded by disabling both models for all customers while its leadership meets with policymakers in Washington to address the issue. The company argues that similar vulnerabilities exist in competing AI systems, yet no comparable restrictions have been applied. Anthropic’s executives are working to clarify the technical basis for the restrictions and advocate for a fairer regulatory approach. Meanwhile, the company has paused public access to the models while negotiations continue. The move comes as AI security remains a growing concern for governments worldwide. The U.S. has increasingly scrutinized advanced AI models, particularly those with potential military or cybersecurity applications. Anthropic’s response highlights tensions between innovation and regulatory oversight in the rapidly evolving AI sector. SpaceX’s record-breaking IPO was also noted in the same report, though unrelated to the AI restrictions. The focus remains on Anthropic’s efforts to resolve the export ban while maintaining transparency with regulators. The outcome of these discussions could set a precedent for how advanced AI models are governed globally.