Anthropic's Safety Superpower

Anthropic released Fable, a new AI model with advanced capabilities that impressed observers but was quickly jailbroken shortly after its public release. Following the discovery of the jailbreak, the U.S. government issued an export control directive suspending access to Fable and its predecessor Mythos for all foreign nationals, citing national security concerns, though Anthropic disputes the severity of the vulnerability and argues that jailbreaks are inevitable and narrow rather than universal.