Anthropic、「Fable 5」の生物学の制限を緩和 誤検知によるフォールバックを約85％削減
ITmedia AI+ja
Anthropicは、AIモデル「Claude Fable 5」の生物学分野における過度な保護機能を緩和したと発表した。安全性を重視するあまり発生していた無害な質問への誤検知や下位モデルへのフォールバックを大幅に削減。専門的な二重用途研究への制限は維持しつつ、一般的な健康・教育用途での利便性を向上させた。
This is a short summary published by AI Global Wire. The full article is owned and hosted by ITmedia AI+ — open it there to read it in full.
Read the full story at ITmedia AI+- Anthropic
Related AI news
- Anthropic announces a feature that allows different Claude Code sessions to message each other with updates and other information, available on macOS and Linux (Marcus Mendes/9to5Mac)Techmeme · August 8, 2026
- OpenAI puts the brakes on a new model because it’s supposedly too powerfulThe Verge AI · August 7, 2026
- Anthropic loosens Fable 5's biology restrictions but keeps the guardrails on for virology and toxicologyThe Decoder · August 7, 2026
- What’s behind the Google AI shakeupThe Verge AI · August 7, 2026
- What’s behind the Google AI shake-upThe Verge AI · August 7, 2026
- Jill Lepore on the ‘Artificial State’ and why Silicon Valley’s leaders are bad sci-fi readersTechCrunch AI · August 7, 2026