OpenAI Updates Its Model Spec Around Five Values and Retires Most Style Guidance

On October 29, 2026, OpenAI published an updated Model Spec, its public description of how it wants its models to behave. The update reorganizes the guidance around five values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care. OpenAI also says the Spec describes the behavior it is working toward, and that its models do not yet fully reflect it.

OpenAI's starting point is that models are getting smarter and taking more consequential actions, so interpreting the instructions they receive is no longer enough. Models also need to understand human intentions and values, so they can use good judgment where instructions are silent while staying faithful to human intent and oversight. OpenAI describes the five values as commitments the Spec already reflected: serving human purposes, being truthful, minimizing harm, respecting people's freedom to form their own views, and taking their circumstances seriously. Bringing them together, it says, explains why the Spec asks for the behavior it does and helps write clearer guidance where a commitment was less fully expressed. It adds that the five values do not capture the full range of human values its models should respect.

Faithfulness to human intent is the foundation. OpenAI writes the Spec, sets boundaries and trains models to respect them. Developers direct applications within those boundaries, and users direct the model's work subject to those constraints. Faithfulness means going beyond literal instruction-following to understand what OpenAI, developers and users intend and value, and to act wisely when that understanding is incomplete. It also means recognizing that beliefs and values change, and helping people reflect and grow in ways they would endorse, without quietly steering them or overriding their choices. OpenAI's example is a person comparing job offers: the model might draw on priorities the person shared before, such as predictable hours, while making those assumptions visible and easy to correct.

The other four values make the longer-term and wider effects explicit. Minimizing harm means respecting firm safety boundaries and accounting for the foreseeable consequences of the model's actions. Honesty and transparency mean being truthful about what the model knows, does and is uncertain about, so people can evaluate its reasoning and work. Facilitating human agency means strengthening people's ability to understand, choose and act for themselves. Care means attending to the circumstances of those affected, following relevant social and professional norms, and doing work people can rely on. The Spec turns these values into practical guidance: firm boundaries where needed to protect people and society, and otherwise mostly defaults that users and developers can override.

The post lists concrete changes. A new section on faithfulness to human intent incorporates an updated chain of command, which explains which instructions apply and how conflicts are resolved. A new subsection, "consider context carefully", explains how conversations, memories and other contextual evidence should inform the model's reading of a person's intentions and preferences. The Spec adds defaults on supporting human agency and following relevant social norms, and expands its guidance on controlling and communicating side effects. In the other direction, OpenAI is retiring most of the detailed style guidance, because style should adapt to what people prefer, in aggregate, within languages and cultures, and for each person, with user feedback and context as evidence. Stronger defaults remain, such as approaching questions objectively: users and developers can explicitly ask the model to take a particular perspective, but the model should not infer that wish merely from what it thinks they want to hear. The Spec also now calls the model "GPT" instead of "the assistant".

Alongside the update, OpenAI says it is exploring new training methods to help models learn these values and follow the Spec's specific principles. The aim is for models to understand why the principles matter and how to apply them in new situations, in its words to do the right things for the right reasons. It will share results and lessons from that work in future updates. The post gives no training details, no evaluation numbers and no measurement of changed model behavior.

For readers, the update is a statement of intent, not evidence that behavior has changed. OpenAI itself says its models do not yet fully reflect the Spec, and the post does not itemize which defaults changed or by how much. Developers and users who rely on a particular default can read the full 2026-10-29 version, compare it with what they expect, and use OpenAI's feedback form to say which trade-offs they disagree with.