OpenAI、AIの行動指針「Model Spec」を改定 “記憶”の使い方や人間による監督を明記
ITmedia AI+ja
OpenAIは、AIの振る舞い指針「Model Spec」の改定版を公開した。全体を5つの価値観に沿って再構成し、メモリや会話履歴の適切な扱い、エージェント自律動作時の人間による監督への服従などを規定。一方で人格設定や詳細な文体規定の大半を削除し、モデル呼称を「the assistant」から「GPT」に改めた。
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+- OpenAI
Related AI news
- Amazon AWS posts 39% revenue growth as cloud rivals also accelerateCNBC Technology · October 29, 2026
- OpenAI doubles usage limits on Pro plans as competition with Anthropic heats upThe Decoder · October 29, 2026
- OpenAI-backed Thrive Holdings acquires ADTTechCrunch AI · October 29, 2026
- OpenAI updates Model Spec to reorganize it around five values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care (OpenAI)Techmeme · October 29, 2026
- An updated Model Spec for more capable and beneficial AIOpenAI · October 29, 2026
- Mistral’s Le Chonk is now a top open model for coding, ranking alongside Codex GPT-6.1 TerraZDNET AI · October 29, 2026