LinkedIn's New "Seems Like AI Slop" Button, Explained
LinkedIn just gave every user a way to report posts they think were written entirely by AI — and people have already clicked it over a million times. This guide covers what triggered the change, exactly what the button does when you use it, and what it actually means if you use AI to help write your own posts.
LinkedIn's new report option sits in the three-dot menu on every post.
LinkedIn Now Lets Users Report AI-Generated Posts
On July 30, 2026, LinkedIn added a "Seems like AI slop" option to the three-dot menu on every post. Selecting it hides that post from your own feed immediately and sends a signal into LinkedIn's detection pipeline, which the company says it uses to tune its classifiers and reduce how far flagged posts spread. LinkedIn's chief product officer, Hari Srinivasan, described the goal as training better models on what users themselves consider low-quality AI content, rather than relying purely on automated detection.
The feature didn't appear in isolation. LinkedIn also scaled back its own AI writing assistant — replacing the "Enhance your post" feature, which could substantially rewrite a draft, with a more conservative proofreading tool built to preserve the original writer's voice rather than generate new content wholesale.
The Research Behind the Button
The trigger was a study from AI-detection company Pangram Labs, published in early July 2026. Using a Chrome extension that scanned more than a million posts people actually scrolled past across LinkedIn, X, Reddit, Substack, and Medium over about two and a half months, Pangram found that 41% of LinkedIn's long-form posts — those over 250 words — were flagged as fully AI-generated. That was the highest rate of any platform in the study, and LinkedIn posts made up roughly a third of everything scanned while accounting for 62% of all flagged AI content across the five platforms combined.
A separate analysis from a competing detection firm, Originality.ai, reached a broadly similar conclusion, reviewing thousands of LinkedIn posts from prominent accounts and finding more than half showed substantial signs of AI use by its own detection threshold. Two different methodologies landing in the same range is part of why the story got traction beyond a single report.
What Happens When a Post Gets Flagged
A user reports the post
Anyone can select "Seems like AI slop" from the three-dot menu on a post. This is a report, not a fact-check — LinkedIn hasn't described any verification step between the click and the next stage.
It disappears from that person's feed
The immediate effect is similar to marking a post "not interested" — it's hidden from the reporter's own feed right away, regardless of what happens next.
The report feeds a detection pipeline
LinkedIn says reports help train and tune its classification models over time, rather than triggering an instant, individual penalty for the post in question.
Reach can shrink for flagged content
Posts that accumulate enough signal, whether from user reports or LinkedIn's own classifiers, see reduced distribution, similar to how other low-quality or spam content gets deprioritized on the platform.
LinkedIn hasn't published the technical details of how its classifiers weigh a single user report against everything else, and the company has been candid that "slop" itself is a moving target it expects to keep refining.
Does This Mean No More AI-Assisted Writing?
Not quite, and LinkedIn's own framing draws a distinction worth paying attention to. The company's target, by its own description, is low-effort, fully machine-generated content — not AI-assisted writing in general, which is a much larger and more normal category on any platform in 2026.
If you use AI to help draft posts, the practical takeaway isn't to stop — it's to make sure the finished post says something you'd actually stand behind: a specific number from your own work, an opinion you'd defend to a colleague, an observation that couldn't have come from anyone else's account. Posts that read as generic, interchangeable, or stitched from common AI phrasing patterns are the ones most likely to get reported, independent of whether AI touched the draft at all.
Bottom Line
LinkedIn added a user-facing report button for AI-generated content on July 30, 2026, after research found that roughly 41% of long-form posts on the platform were fully AI-written — by far the highest share of any major social network measured. The button hides flagged posts from the reporter's feed immediately and feeds a broader detection system that can reduce a post's reach over time. It targets low-effort, fully AI-generated content specifically, not AI-assisted writing as a category, though the detection tools behind it — like all AI detectors — aren't flawless.
More AI Guides, Written Clearly
Browse our full library of AI tool guides, updated regularly as models and platforms change.
Common Questions
What does LinkedIn's "Seems like AI slop" button actually do?
It lets any user flag a post they suspect is AI-generated from the three-dot menu. LinkedIn says the report hides the post from that person's feed immediately and feeds a detection pipeline that can reduce the post's reach for others over time.
Why did LinkedIn add this feature?
AI-detection firm Pangram Labs published research in July 2026 showing that roughly 41% of long-form LinkedIn posts were flagged as fully AI-generated, the highest rate of any platform in its study, and that LinkedIn accounted for the majority of all AI content flagged across five major social networks.
Will using AI to help write a post get it flagged?
Not automatically. LinkedIn has said it's targeting low-effort, fully machine-generated content rather than AI-assisted writing broadly. Posts that read as generic, repeat common AI phrasing patterns, or add nothing the writer couldn't have said themselves are the ones most likely to get flagged.
Are AI detectors on LinkedIn accurate?
Not perfectly. AI detection tools, including the ones behind this research, have documented false-positive rates and can disproportionately misclassify non-native English speakers' writing as AI-generated. LinkedIn hasn't published details on how its own classifiers weigh user reports against that risk.
More from the creator
Explore two other projects from the same creator.