LinkedIn is cracking down on AI slop: how to post with AI and not get flagged

I built an AI tool that drafts LinkedIn posts, so LinkedIn's AI slop crackdown is aimed squarely at people who use products like mine. I think it's the right call. Here's what LinkedIn has actually said, what it counts as slop, and how I'd post with AI now.

What LinkedIn changed

This summer, LinkedIn added a "Seems like AI slop" option to the report menu on posts and comments. Chief Product Officer Hari Srinivasan said the reports help tune LinkedIn's models and that the company is also working to detect slop and other low-quality content itself (Social Media Today). At the same time, LinkedIn removed its own "enhance your post" AI writer and replaced it with a proofreading tool that keeps your voice.

It moved fast. A few weeks later Srinivasan said the option had been used more than a million times and that "members are now experiencing 40% less views on what we classify as AI slop" (Engadget). Authors are told too. When members flag your post, a note appears in its analytics saying some members felt it seemed like AI.

In October, LinkedIn's VP of Engineering Tim Jurka said its classifiers, now covering posts shared beyond your immediate network, detect AI slop with 94% precision (Social Media Today). So this isn't only about who presses the button. LinkedIn is sorting posts itself before they reach people who don't know you.

What LinkedIn counts as slop

LinkedIn's definition is short: "generic and/or repetitive posts that added no real insight" (Social Media Today).

Notice what that doesn't say. It doesn't say "written with AI". Srinivasan said the goal is to show when AI is being overused while still letting people use AI to refine their thinking (Social Media Today). The problem is the post that could have come from anyone: the glossy "5 lessons I learned about leadership" with no lesson you actually learned, the same hook format every day, the list of truisms with an emoji on each line.

How to post with AI and not get flagged

Start from something only you have. That might be a number from your own business, a customer conversation, a mistake you made last week, an opinion people in your field would argue with. If the first line could sit under anyone's name, rewrite it. Flypost's study of 116,000+ posts found that posts about something the writer actually did earned about 33% more reactions, and a concrete detail added about 15% (Flypost).

Take a side. "Insight" in LinkedIn's definition means a viewpoint. Say what you think and why. A post that agrees with everyone adds nothing.

Use AI for the boring parts. That means tightening a draft, suggesting three hooks to pick from, turning a list into carousel slides, fixing the grammar. Those are the uses LinkedIn itself is leaning into.

Don't batch-generate a month of posts from one prompt. That's how you end up repetitive, which is the other half of the definition. Two to five real posts a week beat a daily stream of filler.

Read every post before it goes out, and keep an eye on your analytics. If LinkedIn tells you members flagged a post, work out what made it read as generic and stop doing that.

Where Deynix fits

This is why Deynix works the way it does. With approval on, nothing posts until you've read and signed off on the exact version, so a person is always the last step, not a prompt. Our writer is told to open with a real viewpoint and a specific detail and never to write generic filler, and it works from your brand context (what you sell, who you sell to and how you talk) rather than a blank prompt. On X, reply drafts learn from your own past posts so they sound like you. You can still write the whole thing yourself and just use us to schedule it, which is often the best option.

AI that helps you say what you think faster is fine. AI that says things nobody thinks is what LinkedIn is now filtering out, and honestly, so are your readers. If you want the rest of what's working on LinkedIn right now, my 2026 LinkedIn growth write-up has ten habits with a source behind each one.

Sources

• "LinkedIn offers the option to report AI slop", Social Media Today, 2 August 2026.
• Karissa Bell, "LinkedIn says its AI slop button is working", Engadget, 21 August 2026.
• Andrew Hutchinson, "LinkedIn exec offers insight into the battle against AI slop", Social Media Today, 8 October 2026.
• Sam Cassman, "How does the LinkedIn algorithm work in 2026? 116,000+ posts analysed", Flypost, last verified 12 October 2026.