OpenAI reorganises its Model Spec around five values

OpenAI's post, dated 29 October 2026, says it is updating the Model Spec, which it describes as its public account of how it wants its models to behave. The update organises the guidance around five values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care. The post says the revision adds guidance on context, oversight and agency, adjusts some behavioural defaults, and retires much of the detailed style guidance.

The post says faithfulness to human intent means going beyond literal instructions to what OpenAI, developers and users intend. It says minimising harm means respecting firm safety boundaries and accounting for foreseeable consequences. Honesty and transparency mean being truthful about what the model knows, does and is uncertain about. Human agency means strengthening people's ability to understand, choose and act. Care means attending to the circumstances of those affected and following relevant social and professional norms. The post says the Spec turns these into practical guidance, with firm boundaries where needed, and that otherwise it mostly takes the form of defaults that users and developers can override.

The post says OpenAI is retiring most of the Spec's detailed style guidance, and that style should adapt to what people prefer. It says the Spec keeps stronger defaults, such as approaching questions objectively. It also says the Spec now refers to the model as "GPT" rather than "the assistant". The post says the Model Spec describes the behaviour OpenAI is working toward, and that its models do not yet fully reflect it. It says results from related training work will come in future updates. The post does not announce a new model, a price, or a date when current models will match the revised Spec.

What it means for business: this is a public description of intended model behaviour, not a new product and not a legal duty. Teams that rely on a particular style or refusal pattern may want to read the revised Spec, because OpenAI says current models do not yet fully follow it.