OpenAI updates its Model Spec around five core values

OpenAI published an updated Model Spec, its public description of how it wants its models to behave. The revision reorganizes the guidance around five values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care. Everything below is OpenAI's account.

• New: a section on faithfulness to human intent, with an updated chain of command and a subsection on considering context carefully (conversations, memories and other evidence about what people want). It also adds defaults on supporting human agency and following social norms, and expands guidance on controlling and communicating side effects.
• Retired: most of the detailed style guidance, because OpenAI says style should adapt to what people prefer. Stronger defaults stay, such as approaching questions objectively.
• Wording: the Spec now refers to the model as "GPT" rather than "the assistant".

OpenAI says its models do not yet fully reflect the Spec and that it is exploring new training methods to teach the values, with results to come. It invites public feedback.