Anthropic Launches Claude Fable 5: Mythos-Class AI With Cybersecurity Guardrails 

Anthropic Launches Claude Fable 5: Mythos-Class AI With Cybersecurity Guardrails 
Anthropic has released Claude Fable 5 for general availability, adding strict safeguards that automatically limit its use in high-risk areas like cybersecurity and biology by falling back to Claude Opus 4.8. The company also upgraded trusted Project Glasswing participants to Claude Mythos 5, including partners such as Dragos, Tenable, TrendAI, Netskope, BeyondTrust, Rubrik, BT, Intercontinental Exchange, and Hitachi, after extensive red-teaming found no universal jailbreaks. #Anthropic #ClaudeFable5 #ClaudeMythos5 #ProjectGlasswing #ClaudeOpus4.8

Keypoints

  • Claude Fable 5 is now generally available.
  • The model includes safeguards that restrict high-risk cybersecurity and biology use.
  • Sensitive prompts automatically fall back to Claude Opus 4.8.
  • Anthropic says red-teaming and bug bounty testing found no universal jailbreaks.
  • Trusted Project Glasswing partners were upgraded to Claude Mythos 5.

Read More: https://www.securityweek.com/anthropic-launches-claude-fable-5-mythos-class-ai-with-cybersecurity-guardrails/