Control 1
Rules
Voice, words to avoid, colours, fonts. Short and with examples.
Off-brand AI content is a setup problem. Five simple controls fix most of it.
Free: your brand in 3 real designs, in 48 hours.
To stop AI making off-brand content, use five controls: written rules, approved facts, locked designs, a check before posting, and one owner.
Rules tell AI what to do. Locked designs stop it changing the look. The check and the owner catch what slips through.
Add them one at a time. Even two help.
Control 1
Voice, words to avoid, colours, fonts. Short and with examples.
Control 2
A list AI may quote. Anything else is flagged, not invented.
Control 3
AI changes words, not layouts, logo or colours.
Control 4
Words against voice. Photos against the look. Before posting.
Control 5
One person who updates the rules and handles questions.

Match what you see to the control.
| What goes wrong | Control that fixes it |
|---|---|
| Salesy words you never use | Rules: a never-write list |
| Invented customer names or numbers | Approved facts |
| Wrong logo, odd colours | Locked designs |
| Stock photo in the wrong mood | Check: photo grading |
| Every person sounds different | Owner and shared rules |
The rules part is in brand rules for ChatGPT and Claude teams. Facts and voice: brand voice guidelines.
Free tool
See how an AI tool describes your company today. It is free and runs right here.
Run it before anything AI-made goes out.
1. Voice: does it sound like us? (yes / fix) 2. Words: any word from our never-write list? (no / fix) 3. Facts: every number and name is on the approved list? (yes / remove) 4. Look: logo, colours, fonts are the real ones? (yes / fix) 5. Photo: matches our mood and light? (yes / swap) 6. Claim: could we prove it if asked? (yes / cut) Result: send / fix / stop. Reviewer: [name]
Leave your email. I design them by hand and send them to you within 48 hours. It costs you nothing.
I write to you from [email protected]. If you do not see it, check spam.
Controls people must remember get skipped.
In a Brand OS Portal the controls are built into the tools. Designs are ready and keep their look. Voice Check looks at the words. Image Grader grades the photos. AI Kit gives ChatGPT and Claude your rules. So nothing relies on someone remembering.

Brand OS Pro Portal. The team makes its own posts. No waiting.

Brand OS Core Portal. The CEO writes and designs an on-brand LinkedIn post in about 3 minutes.
A designer is still the right call for the identity itself. See Brand OS Portal and brand compliance.
Tick what is true.
0 of 5 done
Try one Brand OS Portal tool
This is a real tool from a real Brand OS Portal, on the MTG brand. Change the words and the photo. The design stays on brand.
Watch
I made short videos of the tools. Hear me explain them, then scroll through every one at work. Real tools, on a real brand.

It does not know your rules, so it guesses. Give it rules, facts and a design.
Give it a list of approved facts and tell it to flag anything else.
Use ready designs where just words and photos change. A Brand OS Portal works this way.
A trained team member or the brand owner, using a short checklist.
Yes, even two or three help. Start with rules and facts.
Death List Five. Five headaches, one portal. Sound on.
See what ChatGPT says about your company, and how to fix it.
You hear from me within 2 to 8 hours.
May I use cookies for Google Analytics and Clarity to see which pages help? Privacy