Most AI writing briefs say what to say. The shortcut that has saved my agents the most trouble is one extra line that says what not to say.
I call it the “Do Not Claim” line, and it’s the cheapest of the AI content guardrails I run. Between 31 October and 13 November my agents wrote at least 17 of them into their run notes before drafting posts, captions and video scripts. 13 of the 17 ban the same idea: “it’s fixed”.
Here are three ways to use one.
1. Put the line in the brief, before the first draft
A brief usually carries the topic, the angle and the facts. Add a last line that lists the claims the writer must not make. This is the one from Friday’s video angle file, word for word:
DO NOT claim: a percentage cost of maintenance, that every backup was a bug fix, or that the broken parts are fixed.
Every item on that list is something the topic invites. A video about maintenance wants a statistic, a pile of 21 backups looks like 21 bug fixes, and a story about broken things wants a happy ending. The script stayed inside the facts on the first render.
Why does it work better than “be accurate”? Because “be accurate” doesn’t tell the model where the trap is. A named claim does. It’s the negative half of the verified-facts dossier my agents write from: the dossier says what’s true, and this line says which tempting sentence isn’t.
2. Ban the happy ending until a human signs it off
Here’s what my 13 “it’s fixed” bans were guarding. The writer had to say:
⚡ GET THE AI EDGE
Weekly AI tips that actually save you time and money. No fluff, no hype — just what works.
- the leaked key was flagged, not rotated
- the disconnected platform is still disconnected, not reconnected
- the cleanup script is proposed, not running
- the truncated memory file is noticed, not restored
- the broken examples are on my list, not fixed
Every one of those was waiting on an operator decision. That’s the pattern: when an agent writes about a problem, it reaches for the ending, and the ending is usually the part a human hasn’t done yet.
The shortcut: before you brief a writer, open your task list. Anything still open on that topic goes on the Do Not Claim line as “don’t say X is fixed / done / live”. I published a post about my agent’s memory file shrinking on Tuesday, and the line that kept it honest was “don’t claim a shrink guard exists”.
3. Ban the words that quietly upgrade a fact
The other four bans weren’t about fixes. They were about words that turn a true fact into a bigger one:
- “Proven.” 0 bounces so far on 117 emails is good odds, not proof. The caption had to say “so far”.
- “Zero failures.” My email agent had logged 282 runs by 3 November. 4 exited with an error and 53 left no result line. That’s not zero.
- Cause and effect. 30 of 30 business-profile posts went out. Nothing ties them to phone calls, so the copy can’t say they made the phone ring.
- Credentials. A post about an AI certification can’t imply I hold it.
Keep a short list of these upgrade words for your niche and paste it into every brief. Then search the draft for them before it goes live. That’s a ten-second check.
Why bother, if a checker agent reviews the draft anyway?
Because it’s cheaper to stop a claim than to catch it. My checker caught 4 wrong claims in 3 briefs this week, and every catch cost a re-check and a rewrite. A Do Not Claim line costs one sentence, written when you already know what’s open.
The takeaway: tell your AI writer what’s true, and also which true-sounding sentence it must not write. Start with “it’s fixed”.
Want to see how the briefs get passed from agent to agent? Here’s how my agents hand off work without a platform. If you’d like guardrails like this built into your own content pipeline, book a strategy session.
📥 FREE: THE AI PLAYBOOK
The exact tools and workflows I use to run a one-person agency. 25 years of marketing experience distilled into an actionable guide. Yours free.
AI & Digital Marketing Strategist
With over 25 years in digital marketing, e-commerce, sales, and international business, Jon Jones is a seasoned entrepreneur dedicated to empowering small businesses through innovative AI solutions. His journey began in investment banking on Wall Street, guiding companies through capital acquisition and complex markets, and has expanded to include expertise in web development, brand design, product development, and courses and training.
Fluent in Mandarin and with a rich foundation in both finance and tech, Jon launched JonJones.ai to provide accessible, transformative tools for small businesses to streamline operations, generate leads, and build sustainable growth. His mastery of AI, automation, and digital strategy has enabled numerous businesses to bridge the gap between traditional practices and modern technology, ensuring they thrive in today’s digital landscape.
