The AI-Ready Brand Standard

Whether a brand can be reproduced faithfully by someone who does not know it, working with AI.

That is the whole scope. The standard grades the enforceability of a brand's rules: whether each one can be followed without interpretation, and whether a breach can be proved.

What it does not measure

It does not grade whether the brand is any good.

A mediocre brand, reproduced perfectly, is still mediocre. Distinctiveness is not scored anywhere in this standard and cannot be, because distinctiveness is what emerges when well-defined parts are used consistently. It enters in exactly one place: the requirement, in the identity dimension, that positioning is clear enough to rule something out.

It will not tell you the accent should be orange, or which audience the brand should serve. Those are brand decisions. They take taste and a reason you can defend. The standard protects those choices after you have made them.

It also does not mandate a format. The standard judges the enforceability of whatever shape a brand chose. There is no required document structure, no required tool, and no preferred file type. Strict about whether a rule can be executed, agnostic about how the brand got there.

Every rule is graded against both. They apply to any rule in any dimension.

The guessing test

Could AI follow this rule without guessing?

Fails	Passes
"Our palette is warm and energetic."	"The accent covers about 10% of any layout and never sits behind text."
"Photography should feel human."	"Subjects are photographed mid-action, never posed to camera."
"Keep the logo clear of other elements."	"Clear space equals the height of the wordmark's cap height on all four sides."

The violation test

Could you point at a finished piece of work and prove it broke the rule?

Fails	Passes
"Our palette is warm and energetic." Nothing can break it, so nothing can follow it.	"The accent never sits behind text." Any text on the accent proves the breach.

If a rule cannot be broken, it cannot be followed. Rules that survive this test carry a number, a proportion, a boundary, or a never.

If a rule fails either test, it is not a rule. It is a mood. A rule must pass both tests to count as governed.

Every dimension sits at one of five levels. The levels are the unit of the standard; a brand does not have one level, it has a level per dimension.

Level	Name	Definition
0	Absent	Nothing on this dimension
1	Stated	Description and adjectives that a human interprets
2	Specified	Exact values, no instruction about use
3	Governed	Behavior written down: proportions, contexts, never-rules, do and don't pairs
4	Practiced	Governed, plus real approved work with the decisions marked on it

Level 3 is the AI-ready floor. Below it, a tool is guessing. At Level 3, AI can act, and you can prove when it got the rule wrong. Level 4 requires proof: work that already shipped, approved by the people who own the brand. A brand system can be written up to Level 3. Level 4 needs evidence, and it applies selectively (see section 4).

Worked example: color

Level	What the brand has
0	No color section
1	"The color palette is warm and confident, with a bold accent"
2	Blue #1B3A5C, Signal Orange #E8622A, Paper #F8F7F4. Values and names only
3	Signal Orange is for interactive elements. It covers roughly 10% of any layout. It never sits behind text or fills an entire background. On Paper, it clears 4.5:1
4	Level 3, plus an approved layout with the rules marked on it, showing the 10% in use and why the accent sits where it does

Worked example: voice

Level	What the brand has
0	No voice section
1	"We are confident, human and direct"
2	A list of preferred words and a tone adjective per audience
3	Do and don't pairs dense enough to write from, a banned-word list with reasons, rules for when to be loud and when to be quiet, and a before and after rewrite
4	Level 3, plus published copy held up as exemplars, each with the reasoning for why it was written that way

Level 4 is selective by design. Specific examples matter most on the surfaces AI is most likely to generate. A binary rule does not need one. Requiring them everywhere makes a brand system harder to use.

Governed (3) is enough	Needs Practiced (4)
Editorial style	Brand in context
Logo	Layout and composition
Color	Voice and tone
Typography	Photography and imagery
Actions and interactive

These twelve dimensions make up the core of brand guidelines. For each dimension, "Present" is what a completeness check accepts and "Followable" is what AI needs to use it without guessing.

#	Dimension	Present	Followable
1	Identity and strategy	Belief, promise, and audiences included	Promise and belief written in language someone can use. Audiences ranked. Positioning clear enough to rule something out: an enemy, a "we are not."
2	Brand signals	Attributes listed	Each attribute tied to a visible choice. "Rigorous" becomes "structured, grid-based layout." The feeling language stays, because people read the system too, but no attribute is left without a visible choice.
3	Voice and tone	Tone described	Do and don't pairs dense enough to write from, banned words, a before-and-after rewrite, and rules for when the voice gets louder or quieter. Philosophy is welcome, but it never carries a rule on its own.
4	Editorial style	Style notes included	Casing, dashes, numbers, punctuation, links, and attribution decided. The tool should not make a new choice every time.
5	Logo	Files uploaded	Rules for which variant to use and when, where it sits, clear space and minimum size as numbers, what backgrounds it can sit on, and what never to do with it.
6	Color	Palette with hex values	A role for each color, distribution percentages, allowed pairings, contrast floors, and clear limits.
7	Typography	Font families named	Weights matched to the licensed files. A role for each tier. A size scale from display to body, absolute for digital and as ratios, so the hierarchy holds in any medium. Casing and tracking decided.
8	Photography and imagery	Examples uploaded	Treatment written as checkable rules: processing, subjects, what never appears. The examples are described as well as shown.
9	Layout and composition	Grid mentioned	Anchor rules, spacing logic, column limits, and a clear position on whitespace, with numbers where they matter.
10	Brand in context	Example images included	Notes on real work written as rules, not labels. "Content starts top left. The logo aligns with the first text baseline," not "logo." At least one medium shown from start to finish.
11	Actions and interactive	Buttons shown	Exact treatment for each medium, with rules for what can change and what cannot.
12	Anti-patterns	Scattered don'ts	Each don't sits beside the rule it protects, inside its own brand dimension. It earns its place by naming a failure the rule alone does not rule out. Graded on the don'ts at the source, not on a separate list. A system with no don'ts anywhere leaves AI free to improvise.

Ten dimensions calculate the overall score. All twelve determine the floor and tier.

1. Start with each dimension

Score each of the twelve dimensions from 0 to 4. Each point represents one quarter of the scale:

0 = 0% · 1 = 25% · 2 = 50% · 3 = 75% · 4 = 100%

The first goal is to get every core dimension to Level 3. Then reach Level 4 where real examples matter, especially brand in context and at least voice or layout.

The floor is the lowest dimension score. There is no calculation. If voice is at Level 1 and everything else is at Level 3, the floor is 1. Fix that first. A brand usually drifts at its weakest point while everything around it still looks right.

2. Calculate the overall score

The dimensions do not count equally. The weights come from two things: how often a dimension drifts, and how hard that drift is for a tool to catch.

Weighting principle: drift risk ÷ machine enforceability.

Voice can drift in almost every line of copy, and a person still has to judge it. That gives it the highest weight. A logo needs fewer rules, and its worst failure (a wordmark that is typed text rather than an artwork file) can be caught automatically. That gives it the lowest weight. Catastrophic failure in a single dimension is the floor's job, not the weight's.

Weight	Dimensions
15% each	Voice and tone · Color · Layout and composition · Brand in context
10% each	Identity and strategy · Typography
5% each	Brand signals · Logo · Photography and imagery · Actions and interactive

Multiply each dimension's score by its weight, then add the ten results.

For example, a voice score of 3 at 15% contributes 0.45 points:

3 × 0.15 = 0.45

Do the same for the other nine dimensions and add them up. The weights total 100%, so the result stays on the 0-to-4 scale. An overall score of 2.8 is the same as 70%.

Editorial style and anti-patterns are scored but carry no weight of their own. Both already live inside the other ten: editorial style in every line of copy, anti-patterns as the don'ts beside each rule. Weighting them would count them twice. Either can still be the floor, and a low score in either can keep a brand from reaching Governed or AI-ready even when the overall score is high.

3. Determine the tier

The overall score shows how the system is doing as a whole. The individual dimension scores determine which tier it reaches. A strong average cannot cancel a weak floor.

Tier	Requirement
Stated	The brand is described, but the instructions stop at Level 1. A brand book written for people. The PDF without the design intent
Specified	The values exist, but the rules for using them are mostly missing. Most of the system stops at Level 2
Governed	Most core dimensions at Level 3, none below Level 2
AI-ready	Every core dimension at Level 3 or above, and at least two of the four Practiced dimensions at Level 4

Scoring notes

• Half-levels are permitted and in practice necessary. Use one when a dimension genuinely sits on a boundary, and write down why it sits there. A half-level with no stated reason is a rounding error.
• Score the material a tool actually receives, not the material a person can find. A rule that exists only inside a diagram, or only buried in a paragraph, should not be counted as governed until it also exists in a form tools retrieve reliably, such as a labelled table row.
• Score blind where possible, and write expectations down before looking.

Three findings from applying the standard to real brand systems. None of them are requirements; all of them save time.

Rules must include specific context. A rule rarely arrives with the paragraph that introduced it. A tool retrieves it, or someone skims straight to it, and it lands on its own. "Use it sparingly" means nothing by itself. "Use the accent sparingly" leaves no room to guess. Write every rule as though it will be read with nothing around it.

Make it findable. A number inside a diagram depends on the tool being able to read images. A number inside a paragraph is found sometimes and missed sometimes, depending on the tool and the length of the document. Give each number its own label and cell. Retrieval becomes more reliable, and a person can check it faster too.

The fastest route from Specified to Governed is transcription, not invention. Put the people most fluent in the brand in a room with work that is slightly off, and write down every sentence beginning "we would never." Then run each one through the two tests. "We would never make it feel corporate" fails both. "We would never set body copy below 16px" passes both. Most of a brand's governance already exists, unwritten, in the people who have been applying it.