Tips from a CEO ghostwriter to make your writing sound more authentic
Unless you hear an executive say something out loud, odds are the words you’re reading in their latest LinkedIn post or cool op-ed aren’t really theirs. A communications hire like me wrote it.
This kind of support doesn’t come cheap, and if you’re like most founders I meet, you don’t yet have the budget for a dedicated hire or the time to write everything yourself. Luckily, we live in 2026, and AI has put the power of decent writing in your pocket to draft your sales emails, website copy, pitch decks, and—if you’re starting to experiment with visibility—your blogs and LinkedIn posts.
All of that writing compounds into your reputation, because it’s what your target audience sees before they ever meet you. The catch is the “ick” factor—the moment a reader senses that you didn’t write it, the credibility you were building evaporates. They extend that judgment to everything else with your name on it. So the real skill needed is getting AI to write without leaving fingerprints.
The daily digest for entrepreneurs and business leaders
An Inc.com Featured Presentation
If you’re a writer or editor doing this
If you’re a writer or editor doing this work regularly, AI writing tells are pretty obvious, but can easily be missed by the untrained eye. In my ghostwriting practice, where I maintain distinct voices for multiple executives, I keep a running blacklist of obvious AI tropes. These are the most common ones that give your AI usage away that you should strike on sight.
- Em dash: Much to my chagrin, the em dash is the most obvious tell. Many times, it’s used by AI in place of a comma or period. One or two is defensible, but a paragraph built on them is a signature.
- “It’s not this, it’s that”: A negation followed by a swap that sounds deeper than the thing it replaced. “It isn’t a hiring problem, but a clarity problem.” The two halves are rarely in real opposition, so the sentence provides an insight that is not meaningful. Instead, just say what you mean: “It’s a clarity problem.”
- The two-beat aphorism: A short setup sentence followed by a short payoff sentence. “Early chemistry is easy. Long-term trust is hard.” This is the AI’s idea of profundity.
- The standalone punch sentence: This is the dreaded cousin to the two-beat aphorism, and is often presented as a three-word sentence dropped at the end of a paragraph to close it. “That’s real leverage.” Sure, it adds emphasis, but no real information.
- The abstraction: “A successful launch” where the number should be. Generic claims are how AI writes around facts it doesn’t have, and your writing should never be vaguer than your data.
- The comma triad: Everything arrives in threes: “clarity, confidence, and control.” The algorithm loves to round every list up to three because that number feels complete.
- The vocabulary: “Robust.” “Navigate.” “Landscape.” “Unlock.” And don’t even get me started on the adverbs, “quietly” chief among them.
- The bow-tie paragraph: A closing paragraph that summarizes what you just said, like the executive summary of a memo nobody asked for. Instead, end with a call to action or final parting insight.
The biggest tell is the tone. Writing aimed at everyone comes out generic, and “everyone” is exactly who AI writes for by default. The best way to fix tone is by nailing your audience, and that’s most of my job as a communications consultant. Founders hand me the gobbledygook version of their story, and I translate it for whomever needs to hear it. Investors, customers, and the hire you’re trying to close all care about different things.
But your AI can’t know any of that unless you tell it. So name the intended audience every single time you ask it to write. Then put the draft through the two questions every good editor asks: Who is this for, and what can they do with it? An investor update and a customer post should never survive a copy-paste swap.
Deleting bad patterns and identifying your audience only
Deleting bad patterns and identifying your audience only gets you to neutral. To sound like you, give AI actual evidence.
Telling AI to be “warm but authoritative” is like telling a portrait artist you’re “attractive but approachable.” Yeah, you’re hot, but what color is your hair? How tall are you? There are no details included that would make the picture recognizably you. Your voice needs the same sitting a portrait does. Feed the AI real material, like a voice memo, the transcript of your last pitch, emails you wrote before AI existed, the way you actually answer “so what does your company do?” at a dinner party.
Once you’ve assembled your list of banned phrases and example writing, it’s time to give all of it a permanent home. Every major tool has standing instructions that load automatically with each draft instead of living in your head (Skills in Claude, Projects in ChatGPT). Once you get this system in place, the drafts start arriving already sounding like you. So your work shifts from rewriting to directing.
The word directing is the whole game
The word “directing” is the whole game. I use AI in my own practice every day (I affectionately call mine “the robot”). I prompt, it drafts, I direct (sometimes up to 10+ rounds of revisions), and ultimately, nothing ships without my pen. Even before AI, ghostwriting has always worked this way. This level of review and sign-off is what still makes the words ultimately yours.
Get 1 Smart Business Storydelivered straight to your inbox when you subscribe to Inc.’s free daily newsletter.
Source: www.inc.com
