LinkedIn adds a 'Report AI Slop' button as fake-sounding posts take over the feed
Four in ten long LinkedIn posts may already be fully AI-written. Now the platform is letting users flag them and quietly reworking the tools that created the problem in the first place.
Key points
- LinkedIn is adding a button that lets users report any post as "Seems like AI slop," visible in the three-dots menu.
- AI detector Pangram found that 41 percent of long-form LinkedIn posts were flagged as entirely AI-generated.
- Chief product officer Hari Srinivasan called AI slop "a top priority" in a public post announcing the changes.
- LinkedIn is scrapping its AI post-enhancement tool and replacing it with a proofreader that won't rewrite your words.
- New automatic classifiers, software filters that scan posts for signs of machine-writing, will reduce AI-heavy content in suggested feeds.
If your LinkedIn feed has started feeling like a wall of motivational filler written by no one in particular, you're not imagining it. A large share of the platform's longer posts are now entirely written by AI, and LinkedIn is doing something about it.
What is the new button actually doing?
The "Seems like AI slop" option sits inside the three-dots menu on any post. Tap it, LinkedIn hides the post and logs your complaint. According to chief product officer Hari Srinivasan, that feedback will help the company "tune our models and make better feeds."
Behind the scenes, LinkedIn is also deploying new classifiers, automated filters that score posts for signs of heavy AI use or low quality, to cut how often that content surfaces in suggested posts and in content from people you don't follow. Srinivasan also confirmed the company is testing a way to show post authors when other users feel their writing comes across as inauthentic.
We covered Pangram's underlying detection technology on 29 July 2026, when the New York startup raised $9 million and claimed accuracy above 99 percent at spotting AI-generated writing.
How bad has the problem actually gotten?
Bad enough that the numbers are hard to ignore. AI detector Pangram, a tool that analyses text for machine-writing patterns, recently flagged 41 percent of long-form LinkedIn posts as completely AI-generated, a figure first reported by 404Media. That's roughly two in every five longer posts you scroll past containing no words the person who posted it actually typed.
What changes for people who use AI to write their posts?
One specific LinkedIn feature is going away: the option that used AI to "enhance" a post, effectively rewriting what you typed in a smoother, corporate-sounding voice. It's being replaced with a proofreader that fixes spelling and grammar but leaves your words alone.
The intent, Srinivasan says, is to keep your voice yours. Heavy AI rewrites will be harder to do inside the app, though nothing stops someone pasting in text written elsewhere.
What does this mean for ordinary LinkedIn users?
Your feed should, over time, show fewer posts assembled from a template. The report button gives you something concrete to do the next time a post makes your eyes glaze over.
If you post on LinkedIn yourself, the message is plain: write like a person. The platform's systems are now actively trained to spot and downrank content that doesn't sound like one. That's not a small shift for a professional network where a lot of people's careers live.
