# LinkedIn Users Click Its AI Slop Reporting Button a Million Times > Source: > Published: 2026-08-22 07:38:14+00:00 *LinkedIn's new "seems like AI slop" button has been clicked more than a million times in its first two weeks, and the posts it catches are already losing 40% of their views.* Click a post's options menu on LinkedIn now and there's a new choice sitting next to "not interested" and "report": seems like AI slop. LinkedIn rolled the button out on July 30, and Chief Product Officer Hari Srinivasan says members have used it more than a million times since. That's not a quiet feature nobody noticed. That's a crowd of readers rendering a verdict on two years of AI-written posts flooding the feed. The mechanics are blunt. Flag a post as slop, and it gets treated the way LinkedIn already treats content marked "not interested": pushed down, shown to fewer people. Srinivasan says flagged content is now pulling 40% fewer views than it was a few weeks ago. LinkedIn is also rolling out a new notification, tucked into the post analytics dashboard, that tells you privately when your own post has racked up repeated slop reports. No public shaming. Just a quiet nudge that your last "5 lessons I learned from failure" post didn't land the way you hoped. The button didn't appear in a vacuum. ## Where the number comes from In August, Pangram Labs, an AI-detection startup, published research that gave LinkedIn a number it couldn't ignore. The firm ran a Chrome extension that passively scanned more than a million posts real users scrolled past on LinkedIn, X, Reddit, Substack and Medium between April and June 2026. LinkedIn came out worst by a wide margin: more than 40% of its long-form posts, the ones running 250 words or longer, were fully AI-generated. Pangram found LinkedIn accounted for 62% of all the AI-written content it flagged across every platform in the study, despite making up only about a third of the posts it scanned. Comments weren't much better. Roughly 30% of public comments on the platform were fully machine-written, according to Pangram's data. [LinkedIn Launches an AI Slop Button and Kills Its Own AI Writing Tool](https://startupfortune.com/linkedin-launches-an-ai-slop-button-and-kills-its-own-ai-writing-tool/) LinkedIn added a "Seems like AI slop" button to its three-dot menu on July 30, 2026, letting users flag posts for its detection classifiers while it quietly killed its own AI writing tool, enhance your post. The move follows a Pangram Labs study finding 41% of long-form LinkedIn posts were fully AI-generated, the highest rate of any platform... - [linkedin ai generated content detection](https://startupfortune.com/linkedin-launches-an-ai-slop-button-and-kills-its-own-ai-writing-tool/) - [how to report ai spam posts](https://startupfortune.com/linkedin-launches-an-ai-slop-button-and-kills-its-own-ai-writing-tool/) That's the backdrop Srinivasan is working against. It explains why LinkedIn moved on more than one front at once. The company says it has blocked billions of automated comment attempts in recent months, according to Fortune's reporting on the rollout. And it killed off "enhance your post," the built-in AI tool that used to rewrite your drafts for you, swapping it for a proofreading feature that's supposed to fix grammar without flattening your voice into the same cadence everyone else's AI is already producing. ## What it means for founders For founders and operators who treat LinkedIn as a hiring pipeline and a thought-leadership stage, this matters in a very practical way. If you've been using ChatGPT to draft your posts wholesale and publishing them unedited, you're now competing against your own audience's ability to flag you. And that costs you 40% of your reach. That's not a rounding error. That's the difference between a post that reaches your network and one that quietly dies in the feed. Here's the harder question nobody at LinkedIn has answered yet. What happens to the founder who genuinely writes their own posts but runs them through an AI editor first? Srinivasan hasn't said whether the detection LinkedIn is layering under user reports can tell the difference between fully generated slop and a real draft that got polished. Pangram's own methodology flags full generation, not light editing. But LinkedIn hasn't published its own detection criteria, and a crowd of readers clicking a button isn't a detector. It's a vibe check at scale, and vibe checks catch false positives. Frankly, that ambiguity is the real story here, not the million-click headline number. A platform letting its own users decide what counts as slop is cheap to build and fast to ship. But it also means the definition of "AI slop" on LinkedIn right now is whatever a large enough group of people feel like clicking on any given day. LinkedIn gets to point to a 40% drop in views and call it working. Whether it's actually separating machine noise from real, if AI-assisted, writing is a different question, and one the company hasn't shown its work on yet. **Also read:** [Amazon Quietly Raised Prices on Echo, Kindle, Fire TV and Eero Devices](https://startupfortune.com/amazon-quietly-raised-prices-on-echo-kindle-fire-tv-and-eero-devices/) • [Anthropic Puts Its Most Restricted Model Claude Mythos 5 Into Claude Security](https://startupfortune.com/anthropic-puts-its-most-restricted-model-claude-mythos-5-into-claude-security/) • [OpenAI Cuts GPT-5.6 Sol API Prices After Holding the Line for Months](https://startupfortune.com/openai-cuts-gpt-56-sol-api-prices-after-holding-the-line-for-months/)