AI Reports · OpenAI · 2026-10-29 · Policy position
An updated Model Spec for more capable and beneficial AI
OpenAI's updated Model Spec outlines five core values intended to guide more capable AI toward sound judgment, human agency, and faithfulness to human intent.