OpenAI reorganizes Model Spec around five values
OpenAI published a revision of the Model Spec, its public description of the behavior it aims to guide in GPT models.
- The revision highlights five values for guiding GPT behavior.
- Guidance on context, oversight, and human agency has been expanded.
- OpenAI says its models do not yet fully reflect the document.
Context and analysis by Noosigma AI
Editorial summary and context based on the cited source. Claims and results attributed to the organization are reports from that source; they do not constitute independent validation by Noosigma AI.
The new structure groups guidance around five values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care. It also expands guidance on context, oversight, and the effects of model actions.
Why it matters
The revision adjusts response defaults and removes much of the detailed style guidance, which should account for stated user preferences and context. OpenAI says its models do not yet fully reflect the document.
As a public reference, the Model Spec helps explain the company’s principles, but it does not guarantee that every current response will follow all of its guidance.
What to check
Source: OpenAI, October 29, 2026.