How to apply your brand guidelines to AI-generated images

Start with the image's job

“Make an image in our brand style” leaves the subject, lighting and layout undecided. Start with:

• Purpose: audience, message and destination.
• Requirements: dimensions, exact copy and anything the image must not imply.
• Original elements: product, interface or mark that must stay recognisable, with its source file.
• Creative choices: proposed setting, camera angle and space for text.

Keep image-specific choices separate from standing rules: “leave the left third open” applies to this layout; “never stretch the logo” applies across the brand.

Turn the guide into decisions

Select the rules and references that change this picture:

Explain each reference's role: a lighting example does not establish the product or typography. Say where colours belong, not just their values. Where guidance is silent, label the creative choice as a proposal for this image.

Choose inputs the model can use

Check tool support for references, editing, masks and dimensions. Google documents editing and combined references; OpenAI also notes limits in precise text, composition and brand consistency.

Make the input roles explicit for the next agent or designer:

Task and destination:
Dimensions and required crops:
Message and exact copy:
Brand rules relevant to this image:
Reference files and the purpose of each:
Original assets to preserve:
New creative choices for this task:
What to generate:
What to compose or typeset afterward:

Open the referenced files from the tool's environment; naming a file does not give the tool access.

Preserve the parts that must match

Compose original logos, screenshots and exact product details into the final layout. Typeset required copy with the actual font. William & Mary's guidance similarly calls for official identity assets rather than approximations.

For consumer-electronics brand Nothing, Cursor made a Headphone (1) Pro launch preview. Its generated wordmark lost the registered mark, and Rainbrand Review found a product-name spacing error:

Cursor's Agent placed the original wordmark and set the product name in N Type82 over the image:

The revised preview passed colour, typography, logo and copy checks.

Give agents current brand context

Rainbrand maintains brand context across recurring work. Request a Brief describing the image, audience, destination and constraints; it selects relevant guidance and assets.

Pass the completed guidance and files with the task to your creation tools. For a supported model through the AI Gateway, pass the completed Brief with the generation request; naming the brand alone is insufficient.

Request a Brief for [brand] for the image described in the attached task. Use its audience, destination, dimensions and exact copy. Show the relevant image direction and selected assets. Separate references to follow from files to preserve. Propose what to generate and what to place as original artwork before making the image.

The campaign image guide shows how to carry a selected direction into several formats.

Change one thing at a time

1. Name the error: product detail, accent colour, open space or reference use.
2. Preserve what works: “Keep the product view and lighting; remove the decorative objects and leave more space above it.” Fix a wrong product before polishing its background.
3. Adapt the selected direction to each format, then recheck subject, copy and hierarchy. A portrait placement may need a new composition.

Inspect the finished composition

Check the final composition after typesetting, logo placement and cropping, at its intended display size:

Rainbrand Review checks the finished image or capture against the brand. Include expected copy and constraints, inspect its findings, correct with your editing tools and review again.

Keep the checked composition, original files and Brief together for the next format.