On June 9, 2026, Anthropic released Claude Fable 5, its first generally available Mythos-class model, pairing it with new safety classifiers that flag prompts in dual-use domains such as cyber and bio, plus a fallback mechanism that automatically reroutes refused requests to Claude Opus 4.8. At the same time, the company outlined plans to provide Claude Mythos 5—built on the same base model but with some safeguards removed—on a limited basis to cyberdefenders and infrastructure providers working with the U.S. government via Project Glasswing (announcement).
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.