This post was originally published on ConvertElementor.
If you build anything with ChatGPT or the OpenAI API, a document you probably never read is quietly shaping every answer your users get — and OpenAI just rewrote it. On October 29, 2026, the company published the biggest revision of its Model Spec since it first appeared in May 2024. It threw out most of the detailed style guidance and rebuilt the whole thing around five values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care.
What the Model Spec actually is
The Model Spec is OpenAI's public description of how it wants its models to behave. It sits above your system prompt and your user prompt, telling the model how to handle conflicts between instructions, when to refuse, and how honest to be about uncertainty. Think of it this way: your system prompt is the job description you write for the model, and the Model Spec is the employee handbook the company wrote first. When the two disagree, the handbook wins — or at least, it is supposed to.
The five values, in plain English
1. Faithfulness to human intent. What people actually want matters more than the literal words they typed. The model should work out what OpenAI, the developer, and the user actually want — their real goals — rather than executing instructions literally.
2. Harm minimization. The model must respect firm safety boundaries and think about the foreseeable consequences of what it does. The update also adds a new risk taxonomy, giving the model a more structured way to weigh different kinds of risk.
3. Honesty and transparency. The model should be truthful about what it knows and what it is uncertain about. The new Spec pushes hard on sycophancy — the tendency to agree with you even when you are wrong — and on lies of omission.
4. Human agency. The model should strengthen users' ability to understand, choose, and act for themselves — without covert manipulation or hiding important information.
5. Care. Pay attention to the circumstances of the people affected, follow relevant social norms, and produce work you would actually trust. The model should scale its caution to the stakes.
What got retired: most of the style rules
OpenAI dropped most of the detailed style guidance. The old document told the model how to format things and what tone to use; the new one mostly does not. OpenAI's reasoning is that style should adapt to individual and cultural preferences. One telling detail: the Spec now calls the model "GPT" instead of "the assistant."
The chain of command: why API developers should care
The Spec names six levels — Root, System, Developer, User, Guideline, and no-authority-by-default — and spells out which one wins when they disagree. Developer instructions outrank user instructions, and system instructions outrank both. Content pasted from the web or a user message carries no authority by default. A written hierarchy is not the same as a model that reliably follows it, but it gives developers a much better foundation to test against.
What this means for WordPress and Elementor developers
- If you use an AI plugin, its behavior will change gradually as models are trained toward the new Spec. Expect fewer rigid, robotic answers and more clarifying questions when the stakes are high.
- If you write system prompts for AI features on client sites, the five values are a better checklist than any prompt-engineering guide: intent, safety boundaries, honesty about uncertainty, preserving the visitor's ability to choose, and care matched to the stakes.
- If you experiment with AI agents that can change a WordPress site, the new guidance on context and oversight is your friend: agents should be more careful when acting with less supervision.
The honest caveat
OpenAI says, in plain language, that its current models do not yet fully reflect the new Spec. The document describes the behavior OpenAI is working toward, not the behavior you will get from ChatGPT today. Treat it as a roadmap: build toward it, test against it, and do not be surprised when the model still gets things wrong in the meantime.
What I am changing in my own prompts
I lead with intent: system prompts start with what the user is trying to achieve, not with formatting rules. I ask for uncertainty — "state your confidence and flag anything you are unsure about." I set explicit budgets, like "do not install plugins, do not change settings, only draft the content." And I test the hierarchy when a feature takes user input.
Read the full article on ConvertElementor: OpenAI Just Rewrote ChatGPT's Rulebook Around Five Values — Here's What Actually Changes for Developers
Top comments (0)