Back to blog
Problem & SolutionSeptember 24, 2026/Sohaib G

LinkedIn Is Cutting Your Reach by 40% If Your Posts Sound Like AI

LinkedIn is now showing posts that read like AI filler to far fewer people. In August 2026, the company said posts it classifies as "AI slop" get roughly 40% fewer views. If your posts suddenly stopped reaching people, this could be why.

The good news: it's not a ban, it's not permanent, and it's not about whether you used AI. It's about whether your post sounds like nobody was behind it. That part you can fix.

The problem: your posts are reaching fewer people

Maybe you've noticed it already. Posts that used to get a few thousand views now get a few hundred. Comments dried up. Nothing about your posting changed, but the numbers did.

Here's what happened on LinkedIn this summer:

  • July 30, 2026: LinkedIn added a "Seems like AI slop" button to posts. On the same day, it switched on software that spots low-effort, AI-style posts by itself.
  • August 20, 2026: LinkedIn's chief product officer, Hari Srinivasan, said over a million members had used the button, and posts LinkedIn classifies as slop were getting roughly 40% fewer views. (Fortune)

So LinkedIn is actively pushing these posts down. And it doesn't need anyone to click the button. Its own software is doing a lot of the work.

Why it happens

LinkedIn is full of AI-written posts. One AI detection company, Pangram, found more than 40% of long-form LinkedIn posts it checked were fully AI-written. Members got tired of scrolling past them.

So LinkedIn now looks for posts that seem polished but empty: no real experience, no real point of view, nothing a real person clearly added. (ContentGrip)

Here's the part most people get wrong. LinkedIn has said using AI to improve your writing is fine. Fixing grammar or tightening a draft won't get you punished. What gets punished is a post that could have been written by anyone, about anything, for no one in particular.

The posts most likely to get hit:

  • Generic tip lists ("7 things top performers do every morning").
  • Stories that feel invented, with a perfect lesson at the end.
  • Posts full of words like "leverage," "unlock," and "game-changer."
  • The standard AI shape: one-line hook, three short points, "What do you think?"
  • News summaries with nothing new added.

How to tell if you got hit

A drop in reach isn't always the slop filter. Reach moves around for lots of reasons. Before you change everything, check these:

CheckWhat it tells you
Did every post drop, or just some?If only your AI-drafted posts dropped, that's a strong sign. If everything dropped, look at other causes too.
Do your personal-story posts still do okay?If posts with real details still reach people, the filter is likely hitting your generic ones.
Did you add outside links?Posts with links in them often reach fewer people. Try the link in a comment.
Did you change when or how often you post?Timing changes can move your numbers.
Did LinkedIn send you a notice?LinkedIn has said it's rolling out notices to people whose posts get many slop reports.

You can see each post's views in its analytics on LinkedIn. Compare your last ten posts. Put the ones that did badly next to the ones that did fine. Usually the pattern is obvious once they're side by side.

Possible fixes

There are three ways people respond to this. Only one works well.

1. Stop using AI completely. It works, but it's more than you need. LinkedIn isn't punishing the tool. You'll lose the time AI saves you without gaining much.

2. Run AI posts through a "humanizer." This helps with wording, but it misses the real issue. If the post has no real idea or experience in it, changing the words won't give it any. It's still empty.

3. Change what goes in, then edit what comes out. This is the one that works. Start every post from something real you have, and edit the draft so it sounds like you. The rest of this guide shows how.

The recovery plan

Follow this for your next ten posts:

  1. Start from your own material. A voice memo, notes from a call, a comment you wrote, a video you made. Never a one-line prompt like "write a post about leadership."
  2. Put one real detail in every post. A number from your work. A name. A mistake. Something that happened on a specific day.
  3. Make one point per post. Generic posts try to cover everything. Real posts say one thing.
  4. Cut the AI tells. Clichéd openers, big corporate words, lists of three, em dashes, and "What do you think?" endings.
  5. Read it out loud. If you wouldn't say it to a colleague, rewrite it.
  6. Post less, if you need to. Three strong posts a week beat seven generic ones. Especially now.

There's no official reset button. Each new post is judged on its own, so better posts are the recovery.

Tools that help

Any AI chat tool can still be part of this. The difference is what you feed it. Paste in your own notes or transcript and a few of your old posts so it can match your style. Don't ask it to invent the post from nothing.

Some tools are built around this. Reshaper AI, for example, only drafts from something you already made, like a YouTube video, a transcript, or your raw thoughts, writes in your saved voice, and checks every draft for common AI patterns before you see it. It still needs you to add your own details and read the final version. No tool can put your real experience into a post for you.

For the full method, see how to use AI for LinkedIn posts without sounding like AI. For more on the button itself, see LinkedIn's "Seems like AI slop" button, explained.

Example: the same idea, before and after

A post likely to get pushed down:

In today's fast-paced world, we often overthink our content. But here's the truth: it's not about perfection. It's about authenticity. The posts that perform best are raw, real, and relatable. So stop overthinking and start posting. What's the best post you almost didn't share? Let me know below! 👇

The same idea, rebuilt from a real experience:

I posted every weekday in January for 21 days.

My best one got 9,000 views and I wrote it in 5 mins

I mean it wasn't the one I spent the most time on. It was just a 60-word post about a client who paid me four months late.

This one was something I wrote on the go and I almost didn't post it

The first one could be anyone's. The second one could only be yours. That's the whole difference, and it's exactly what LinkedIn's filter is looking for.

Key takeaways

  • LinkedIn is showing posts it sees as AI slop to roughly 40% fewer people.
  • It's aimed at empty, generic posts, not at every post that used AI.
  • Check your analytics before assuming the filter is the reason for a drop.
  • Humanizer tools alone won't fix a post with nothing real in it.
  • Start from your own material, add one real detail, and read it out loud before you post.

Frequently asked questions

Is LinkedIn punishing everyone who uses AI?

No. LinkedIn has said that using AI to improve your writing is not the problem. The reach cut targets posts it classifies as slop: polished but empty content with no real experience or point of view.

Is the 40% reach cut permanent?

LinkedIn hasn't described it as permanent. It applies to posts LinkedIn classifies as slop, so the way to recover is to publish posts that don't read that way.

Can I get an AI slop label removed from my post?

LinkedIn hasn't announced an appeal process. The button isn't a policy report, so your post won't be taken down. It just gets shown to fewer people.

How long does it take to recover reach?

LinkedIn hasn't given a timeline. In practice, your next posts are judged on their own, so better posts start helping right away. Give it a few weeks of consistent, specific posts before judging.

Try Reshaper AI free

Turn one piece of content into posts for every platform in seconds.

Start free