Anthropic Releases Claude Fable 5, Its Most Powerful AI Yet, With Cyber Safeguards

Anthropic released Claude Fable 5 on June 9, making its most capable AI model generally available. The company took an unusual approach by shipping one underlying model as two distinct products, differentiated by a layer of safety classifiers. Claude Fable 5 is accessible to the public, while its twin, Claude Mythos 5, which has enhanced cyber safeguards removed, is restricted to a vetted group of cybersecurity researchers. This dual release strategy aims to balance broad accessibility with rigorous security testing. The company stated that Mythos 5 is intended for advanced security research and red-teaming efforts. This marks a significant step in Anthropic's development of powerful AI systems while prioritizing safety and responsible deployment. The underlying model powering both Fable 5 and Mythos 5 represents a substantial advancement in Anthropic's AI capabilities, though specific benchmark improvements were not detailed in the announcement.