Anthropic has significantly loosened Claude Fable 5’s biology safeguards cutting biology-related fallbacks by about 85% while keeping higher-risk research restricted.
At launch, Anthropic intentionally used a very broad safety classifier because Fable 5 can outperform experts on some complex biology tasks. The downside was that many harmless questions were unnecessarily redirected to the less capable Opus 5.
The updated classifier is designed to distinguish benign requests from potentially dangerous dual-use work more precisely.
Users should now get Fable 5 more often for:
• Understanding symptoms
• Interpreting lab results
• Educational biology questions
• Clinical tasks for healthcare professionals
Anthropic expects the change to reduce all fallbacks by roughly:
• 67% on Claude ai
• 55% on Cowork
• 17% on Claude Code
• 7% on the Claude Platform
But Fable 5 will still fall back to Opus 5 for sensitive areas including virology, toxicology, molecular design, and other dual-use professional biology research.