How to Make AI Content That Doesn't Sound Like AI
Anti-AI writing rules are a set of explicit instructions given to an AI system that prevent it from producing the recognizable patterns, phrases, and formatting habits that make AI-generated content immediately identifiable. These rules work by telling the AI what NOT to do, which is often more effective than telling it what to do. When combined with a brand voice document and ideal client profile, anti-AI writing rules transform generic AI output into content that reads as if a human wrote it, because a human directed every constraint.
If you've been scrolling Instagram or LinkedIn lately, you've probably noticed something: a lot of orthodontic practice content has started sounding exactly the same. The hooks are identical. The formatting is predictable. The transitions are eerily similar. And every other post seems to end with "and honestly?" followed by some vaguely inspirational conclusion.
That's because most of it is being generated by AI with zero guardrails. Practices are typing "write me a post about phase one treatment" into ChatGPT and publishing whatever comes back. The content is technically correct, grammatically polished, and completely forgettable. It doesn't sound like any specific practice. It sounds like AI.
This is a solvable problem, and I've spent the last year solving it. Here are the specific writing rules that make AI content indistinguishable from human-written content.
Why Most Practices Don't Have Complete SOPs (And Why It Costs Them)
The perceived time commitment is the biggest barrier. Practice managers and team leaders know that sitting down to write out a detailed protocol for, say, how to process an insurance predetermination feels like it's going to take an hour. Multiply that by the dozens of protocols your practice runs on, and the project feels impossible alongside the daily demands of actually running the practice.
But the cost of not having SOPs is real and compounding. Every time a team member leaves (and turnover in orthodontics is a well-documented challenge), institutional knowledge walks out the door. New hires take months to get up to speed because there's no structured training path. Team members interrupt each other constantly with questions about how things are done. And inconsistency creeps in across every process because there's no defined standard.
The practices that have clean, current SOPs didn't create them all at once in a heroic weekend effort. They built them incrementally, and now AI makes that incremental approach dramatically faster.
The 8 Anti-AI Writing Rules
Rule 1: Never use em dashes
This is the single biggest giveaway of AI-generated content. Em dashes (the long dash that looks like this: —) appear in virtually every AI output. Claude loves them. ChatGPT loves them. Human writers use them occasionally. AI uses them constantly. Replace every em dash with a comma, a period, or parentheses. This one change alone makes content feel 30% more human.
Rule 2: No antithesis formatting
AI loves to place two opposite ideas side by side for dramatic effect. "You don't grow by fitting in. You grow by standing out." "It's not about working harder. It's about working smarter." This structure has become so overused in AI content that it now signals "a computer wrote this" to any attentive reader. Just say what you mean directly without the rhetorical mirror.
Rule 3: No stacked short sentences
AI frequently stacks two or three short sentences back to back for emphasis. "AI can help. It can sharpen your words. It can save time." This reads like a robot stacking blocks. Combine related ideas into flowing sentences: "AI helps sharpen your words and save time." One sentence says the same thing with actual rhythm.
Rule 4: Ban the filler phrases
There is a specific set of phrases that AI reaches for when it needs a transition or emphasis. Once you see them, you can't unsee them: "Here's the thing." "The truth is." "At the end of the day." "Let that sink in." "More than ever." "And honestly?" These phrases do zero work in a sentence. They're verbal filler that AI uses as structural crutches. Cut all of them.
Rule 5: No self-answering rhetorical questions
"The systems themselves? They're powerful." "The real question? It's simpler than you think." This pattern (question fragment followed immediately by the answer) is a signature AI move. It creates artificial drama around information that should just be stated. If the systems are powerful, say "The systems are powerful" and keep moving.
Rule 6: No hypothetical role lists
AI loves to demonstrate value by listing use cases by role: "Your TC uses it for scripts. Your FC uses it for claims. Your marketing coordinator uses it for social posts." This is lazy content that reads like a product specification sheet. Instead of listing what each role does with the tool, describe the cultural or operational shift it creates. Show the transformation, not the feature list.
Rule 7: Never fabricate data
This one matters especially for orthodontic content. AI will happily generate fake statistics to strengthen a point. "Studies show that 73% of patients prefer..." If no study was cited, the statistic doesn't exist. Use directional language ("many practices find that...") or pull real data from verified sources. Making up numbers destroys credibility and can create real professional liability in healthcare content.
Rule 8: Apply the rule of three
When listing items, use one or three. Never two. Two items feel incomplete. Three feels complete and rhythmic. One feels deliberate and focused. Two just feels like you ran out of ideas. This is a small rule that has an outsized impact on how professional your content reads.
How to Implement These Rules in Your AI System
The most effective approach is to build these rules into your AI's operating instructions so you never have to remember them manually. In the AI Practice Advantage program, we create what we call a Brand Champion project with custom instructions that include your brand voice, your ideal patient profile, and these anti-AI writing rules built directly into the system.
Once they're in the custom instructions, every piece of content your Brand Champion produces automatically follows these rules. You don't have to prompt it each time. It just knows.
You can also create a separate anti-AI writing rules document and upload it to the knowledge base of any project where you're generating content. This approach works across multiple projects without duplicating the instructions.
The Test: Could a Human Have Written This?
After generating any piece of content, read it with one question in mind: could a smart, experienced human have written this without AI? If the answer is no, if it's too polished, too symmetrical, too perfectly structured, rewrite until the answer is yes.
The irony of great AI content is that it should feel slightly imperfect. Real human writing has rhythm variations, unexpected word choices, and sentences that don't all sound the same length. That "roughness" is what makes it feel authentic.
Frequently Asked Questions
Won't these rules limit what AI can create?
The opposite. Constraints produce better creative output. By eliminating the patterns that make everything sound the same, you force the AI to find more interesting ways to communicate your ideas. The content becomes more distinctive, not less.
Do these rules work with ChatGPT or just Claude?
These rules work with any AI tool. You can add them to ChatGPT custom instructions, Claude project instructions, or any system prompt. The patterns they address are universal across all large language models.
How do I know if my content sounds like AI?
Read it out loud. If you wouldn't naturally say any phrase in the piece, that phrase needs to be rewritten. Also, show it to someone who knows your voice and ask: "Does this sound like me?" Their reaction will tell you immediately.
Want to build an AI content system that sounds like your practice, not like every other practice? The AI Practice Advantage program includes the full Brand Champion build with anti-AI writing rules baked in. Learn more about AIPA.
About the Author: Lindsay Quinn is the CEO and Founder of Heartwise Collective, an orthodontic consulting firm specializing in financial systems audits, accounts receivable recovery, AI implementation, and fractional COO services. She has trained 46+ orthodontic practices in building HIPAA-compliant AI systems through her AI Practice Advantage program and has over 22 years of orthodontic industry experience.