Relevant NewsDaily Edition
Anthropic has taken its Claude Fable 5 model offline in response to a US government order, citing discovery of a method to bypass the AI system's safety guardrails. The company disclosed the jailbreak vulnerability in a blog post but did not specify which government agency issued the directive or the nature of the exploit.
Why it matters: This case sets a potential precedent for government intervention in AI safety issues and raises questions about how companies balance transparency, security protocols, and regulatory compliance in the emerging AI governance landscape.
