How to use AI to write LinkedIn posts that sound like you

Using AI to write LinkedIn posts is completely fine. Sounding like AI is the problem. The flat, over-polished, says-nothing voice that comes out of a one-line prompt is what gets scrolled past (and quietly judged). The fix isn’t to avoid AI. It’s to feed it your voice, prompt it for a specific angle, and edit it like a human before it goes anywhere near publish.

Done well, AI turns a two-hour writing session into ten minutes. Done badly, it fills the feed with posts that all sound the same. Here’s the difference.

Why most AI LinkedIn posts flop

AI writes the average of everything it’s read, and the average is forgettable. Ask a general model for “a LinkedIn post about leadership” and you’ll get competent, confident, completely generic prose that could have come from anyone. On a platform where attention goes to a clear point of view, average is invisible.

The tells are easy to spot once you know them: openers like “In today’s fast-paced world,” phrases like “delve into” and “navigate the landscape,” a tidy list of advice everyone already agrees with, and not one specific detail that proves a real person was there. Readers may not name it, but they feel it, and they keep scrolling.

Step 1: Feed it your voice, not just a topic

The single biggest lever is giving the model examples of how you write. Paste in three to five of your best-performing posts and tell it: match my sentence length, my vocabulary, my rhythm, and how blunt or warm I am. Then write in that voice. “Write in my voice” alone does nothing; voice comes from examples, not instructions.

This is the step almost everyone skips, and it’s why their output sounds generic. It’s also the exact problem a trained brand-voice engine solves automatically. More on that below.

Step 2: Prompt for a specific angle, not “a post”

“Write a LinkedIn post about X” gets you a generic post about X. Give it the idea, the angle, the audience and the format instead. Compare these two prompts:

• Weak: “Write a LinkedIn post about hiring.”
• Strong: “Write a short, punchy LinkedIn post for early-stage founders. Contrarian angle: most first hires fail because of a vague brief, not the candidate. Open with a one-line story hook. One clear opinion, no listicle.”

The second gives the model something to actually say. The more you decide before you prompt, the less generic the draft. If you’re not sure what the angle should be, our guide on what to post on LinkedIn has twelve you can borrow.

Step 3: Edit it like a human (three fixes)

Never post an AI draft raw. Three quick edits close most of the gap:

1. Cut the throat-clearing. Delete the warm-up intro and start on the hook. The first line has to earn the second.
2. Add one thing only you know. Replace a generic claim with a specific number, name or moment from your actual experience. Specifics are what AI can’t invent and readers can’t ignore.
3. Read it aloud and cut anything you’d never say. Then add one line of genuine personality. If it sounds like you talking, you’re done.

Does LinkedIn penalise AI content?

No, LinkedIn doesn’t penalise AI-assisted posts. It penalises boring ones. Reach is decided by whether people stop, read and engage, and your audience is the real filter. A thoughtful, specific post drafted with AI will always beat a lazy one typed by hand. Worry about being generic, not about being caught.

This is the whole reason WordPush trains on your past LinkedIn posts before it writes a word: the first draft starts in your voice and on an idea pulled from your industry, so you’re editing something that already sounds like you, not fighting ChatGPT’s default tone every time. If you’re weighing tools, see how it compares in WordPush vs Taplio or the honest rundown of the best LinkedIn content tools. The tool matters less than the habit: draft fast, edit like a human, and never publish the generic version.