In brief
Anthropic halted the use of its new Claude Fable 5 and Mythos 5 models for foreign users after the US Department of Commerce issued a security order citing "jailbreaking" concerns. The move follows a lawsuit over the company’s tools being deemed a supply‑chain risk by Pentagon officials. Experts warn that the restriction could slow AI development and international cooperation, while the legal battle continues to shape the future of cloud‑based AI services.
On the day the Claude Fable 5 model went live to the public, American authorities ordered Anthropic to halt all use of the tool for foreign nationals, citing uncontrolled security risks stemming from a reported “jailbreak” method that bypasses safety checks. The United States Commerce Department’s directive forced the company to instantly disable Fable 5 and Mythos 5 for its international customer base, an action detailed in a statement posted on Anthropic’s website that morning.
The decision comes amid an ongoing lawsuit where Anthropic has sued the Pentagon over a designation that classifies its AI services as a supply‑chain risk, making them unavailable to U.S. government agencies. A federal judge has ruled that the Pentagon’s order cannot be enforced, allowing U.S. contractors and agencies to keep using Anthropic’s tools while the case proceeds.
Security experts such as Gina Neff of Queen Mary University of London caution that restricting model access could hinder research, testing and safe‑deployment of cutting‑edge AI capabilities. Neff highlighted that the UK AI Security Institute found Anthropic’s model able to bypass network defenses 73% of the time, signifying a dramatic leap in potential cyber‑attack capabilities.
The incident has sparked a broader debate about how quickly AI architectures are being rolled out versus the safeguards required to prevent misuse. While Anthropic claims its pre‑release libraries were for a limited number of vetted organisations to test security boundaries, critics argue that the company’s own warnings about the model being “too powerful” may have been more hype than caution.
In a related story, finance ministers and banking leaders independently raised concerns over the Mythos AI model’s potential to outstrip existing protective measures, illustrating the international uncertainty and regulatory pressure on fast‑growing AI firms.
At 54 minutes ago, active discussions on social media and the PulseWire network show increasing user engagement on the implications for future AI policy and the balance between open innovation and national security. Live updates will continue as court rulings and government responses develop.
