Anthropic loosens Claude's biology filter: about 85% fewer refusals on health questions
Anthropic rewrote the rulebook behind the classifier that decides which biology questions Claude Fable 5 will answer, carved out explicit exceptions for benign uses, then retrained it. Biology-related fallbacks — where the model quietly hands the question to the weaker Opus 5 — fell by roughly 85%. Overall fallbacks dropped about 67% on Claude.ai, 55% on Cowork, 17% on Claude Code and 7% on the platform. Professional virology, toxicology and drug-molecule design stay blocked.
Why it mattersOver-blocking is the tax users pay for safety filters, and labs almost never publish how heavy that tax was. If you have been brushed off asking Claude to explain a lab result or a symptom, that should now mostly stop, while the dual-use line stays where it was. It is also a concrete answer to a complaint that has followed Fable 5 since launch: that the safest model on the market would not talk about cancer.
Read in the app — free, in 9 languages
Related stories
Machine learning read the shape of sick brain cells and picked out nine already-approved drugs that calmed them down2026-08-21Apple Music will tell you when a song was made by a machine — but the uploader decides whether to say so
2026-08-21Stripe has just paid $7.5 billion for a model router. Days later Ramp built one and is giving it away until January.
2026-08-21Meta's assistant is now a Mac app that reads your screen and types into any window — and what it sees can train the model
2026-08-21One click on a news site now tells Google to show you more of it — and you stay on the page you were reading
2026-08-21