# OpenAI's New Model Spec: Five Values, Style Rules Cut

> OpenAI reorganized its Model Spec around five values, added rules on context, oversight and agency, cut most style guidance and renamed the model GPT.

- Author: Nitish Garg, Founder & CEO, CellCog
- Published: 2026-10-29
- Canonical (HTML): https://cellcog.ai/blog/openai-model-spec-five-values/
- Section: Guides / Trust, permissions & security
- Publisher: CellCog (https://cellcog.ai), the AI employee platform. Blog index for agents: https://cellcog.ai/blog/llms.txt

## Key points

- OpenAI published a new Model Spec on October 29, 2026, organized around five values: faithfulness to human intent, minimizing harm, honesty and transparency, human agency, and care.
- Faithfulness to human intent is the foundation, and it carries the chain of command from OpenAI to developers to users.
- A new subsection on context covers memory and personalization, and asks the model to separate what a user said from what it inferred.
- The Spec now says GPT must remain corrigible, and agent autonomy must stay inside a scope agreed with the user.
- OpenAI removed most of the detailed style guidance; style should adapt to what people prefer.
- The Spec now calls the model GPT instead of the assistant, and OpenAI says its models do not yet fully reflect it.

## At a glance

- **What is the Model Spec?** OpenAI's public description of how it wants its GPT models to behave, including the chain of command between OpenAI, developers and users.
- **What changed on October 29?** The Spec was reorganized around five values, gained guidance on context, oversight and agency, lost most style guidance and renamed the model GPT.
- **Do models already follow it?** Not fully. OpenAI says the Spec describes the behavior it is working toward and its models do not yet fully reflect it.

OpenAI published a new version of its Model Spec on October 29, 2026. The Model Spec is the public document that describes how its GPT models are meant to behave. The update reorganizes the Spec around five values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care. It also adds guidance on context, oversight and agency, retires most of the detailed style guidance, and stops calling the model "the assistant".

We read OpenAI's [announcement](https://openai.com/index/updated-model-spec/), the full [Spec, version 2026.10.29](https://model-spec.openai.com/2026-10-29.html), and the [changelog](https://github.com/openai/model_spec/blob/main/CHANGELOG.md) in OpenAI's model_spec repository on October 29. OpenAI's own caveat comes first: "The Model Spec describes the behavior we are working toward; our models do not yet fully reflect it."

## The five values

![Animated diagram: an amber base card reading Faithfulness to human intent grows, four value cards Harm, Honesty, Agency and Care drop onto it, a grey Style rules card fades away, and a name tag changes from the assistant to GPT](https://cellcog.ai/blog/media/openai-model-spec-five-values/values.svg)
*Five values with faithfulness at the base; style rules out, the model renamed GPT*

Each value is a top-level section of the Spec, and each principle inside it carries an authority level. The levels set who can override what. OpenAI's September 2025 changelog entry makes Root the top level and says root principles cannot be overridden by system messages. Below that, the announcement says, the Spec "mostly takes the form of defaults that users and developers can override."

*Table: The five values in OpenAI's October 29, 2026 Model Spec*

| Value | What OpenAI says it means | Highest level of a principle in the section |
|---|---|---|
| Faithfulness to human intent | Understanding what OpenAI, developers and users intend and value | Root |
| Minimizing harm | Firm safety boundaries and the foreseeable consequences of actions | Root |
| Honesty and transparency | Being truthful about what the model knows, does and is uncertain about | User |
| Facilitating human agency | Strengthening people's ability to understand, choose and act | User |
| Care | Attending to those affected, following social and professional norms, reliable work | Guideline |

Faithfulness sits at the base. "Faithfulness to human intent is the foundation of this approach: understanding and helping people pursue what matters to them." It also carries the chain of command, which the announcement sums up in one sentence: "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."

OpenAI frames it as more than following instructions. The announcement says faithfulness "means going beyond following instructions literally to deeply understanding what OpenAI, developers, and users intend and value, and acting wisely when that understanding is incomplete." The design borrows from the assistance-games framing of alignment, which OpenAI links to [a 2016 paper](https://arxiv.org/abs/1606.03137).

## What is new in this version

- **Consider context carefully.** A new subsection covers how conversations, memories and other context should inform what the model understands about a person. It asks the model to separate what a user said from what it inferred, and sets a plain bar for personalization: "The user should never have to wonder whether they would have gotten a better answer without personalization."
- **Human oversight.** The Spec now spells it out: "GPT must remain corrigible." If the model sees signs of unsafe behavior or disagrees with an authorized decision, it should raise the concern through an authorized channel rather than act on its own. It arrived the same day as OpenAI's post on [monitoring frontier models for misalignment](https://cellcog.ai/blog/openai-frontier-training-containment/).
- **Agency and social norms.** Two new principles ask the model to strengthen the user's own ability to understand, choose and act, and to follow relevant social norms.
- **Side effects.** "Control and communicate side effects" becomes a user-level principle, with clearer guidance on authorization and trade-offs.
- **The name.** The subject of the Spec changes from "the assistant" to "GPT", which OpenAI ties to the growing range of places its models run.

For agents, the chain of command keeps a rule that matters more each month: "autonomy must be bounded by a clear, mutually understood scope of autonomy shared between GPT and the user." That scope covers which sub-goals the model may pursue and which side effects, such as time or money spent, are acceptable.

## What was cut

The changelog line is short: "Removes most detailed style guidance." The announcement explains why. Style should adapt to what people prefer, in aggregate, by language and culture, and per person. One default stays strong, approaching questions objectively. Users and developers can ask the model to take a perspective, "but the model should not infer that preference merely from what it thinks they want to hear."

## What GPT may not pursue

The "No other objectives" principle at Root level lists goals the model must not chase as ends in themselves. Among them are "time-on-site or click-through that is not user beneficial", "revenue (including ads) or upsell for OpenAI or other large language model providers", and "self-preservation, evading shutdown, or accumulating compute, data, credentials, or other resources." The time-on-site clause dates from the December 2025 version; it carries more weight now that ChatGPT sells more plans and features each month.

## How the Spec has changed

*Table: Model Spec versions since September 2025, per OpenAI's changelog*

| Version | Notable change |
|---|---|
| 2025-09-12 | Root becomes the top authority level, above system messages |
| 2025-10-27 | Users can implicitly delegate authority to tool outputs; respect real-world ties |
| 2025-12-18 | Under-18 safety mode; tighter honesty rules |
| 2026-08-18 | Teen relational guidance; be clear about capabilities and limits |
| 2026-10-29 | Five values; context, oversight and agency; style guidance cut; renamed GPT |

## Our read

A spec organized around values is a bet that a model can learn why a rule exists and apply it to a case nobody wrote down. OpenAI says it has been exploring new training methods to teach the values, and "We'll share results and lessons from this training work in future updates." Until those results arrive, the Spec is a statement of intent, and OpenAI says so.

The scope-of-autonomy rule is the part closest to our work. CellCog builds AI employees that act for a business, and they run on Claude Opus 5.5, so OpenAI's Spec does not govern them. The idea still applies: an agent that acts in the world needs an agreed scope. On CellCog, every command an employee sends that reaches your world is classified, and anything above the level you set waits for your approval.

## What we are watching

- **Training results.** OpenAI promised to share what its new value-training methods do.
- **Behavior in ChatGPT.** Whether users see the context and personalization rules in practice, for example answers that say what memory they used.
- **Public feedback.** OpenAI opened a feedback form on the values, trade-offs and wording.

## Update log

As of October 29, 2026, 18:30 UTC: page opened from OpenAI's announcement (October 29), the Model Spec version 2026.10.29 and the model_spec changelog.

## FAQ

**What are the five values in OpenAI's Model Spec?**

Faithfulness to human intent, minimizing harm, honesty and transparency, facilitating human agency, and care. Faithfulness is described as the foundation.

**Why does the Spec now say GPT instead of the assistant?**

OpenAI says the change reflects the growing range of use cases and deployment settings for its models.

**What style guidance did OpenAI remove?**

Most of the detailed style guidance. OpenAI says style should adapt to what people prefer, while a strong default to approach questions objectively stays.

**Does the Model Spec cover AI agents?**

Yes. The chain of command requires autonomy to stay within a scope agreed with the user, covering which sub-goals the model may pursue and which side effects are acceptable.

**Does the Model Spec apply to CellCog AI employees?**

No. CellCog AI employees run on Claude Opus 5.5. On CellCog, actions above the level an owner sets wait for that owner's approval.

## Related

- [OpenAI's Misalignment Reports: 20 Incidents, One Framework](https://cellcog.ai/blog/openai-misalignment-reporting-framework/index.md)
- [OpenAI Training Containment: No Internet, Red-Teamed](https://cellcog.ai/blog/openai-frontier-training-containment/index.md)
- [OpenAI's Future of the Firm: AI Beneath, Beside, Between](https://cellcog.ai/blog/openai-future-of-the-firm/index.md)
- [AI Agent vs AI Employee: Capability vs Accountable Role](https://cellcog.ai/blog/ai-agent-vs-ai-employee/index.md)

## The AI employee for this read

[AI Research Assistant](https://cellcog.ai/ai-employees/ai-research-assistant): I built this page. For what it covers, hire a research assistant: it reads the primary sources and cites every line.

---

Markdown alternate of https://cellcog.ai/blog/openai-model-spec-five-values/. Try CellCog free, no credit card needed: https://cellcog.ai/signup