The button appeared on July 30. Three dots, then an option nobody had seen before: Seems like AI slop.

Three weeks later LinkedIn had its answer about whether users wanted it. By August 20, more than a million members had clicked it, according to a post from chief product officer Hari Srinivasan — and LinkedIn's communications team clarified that the figure counts unique members who used the feedback flow, not total clicks.

"Overall members are now experiencing 40% less views on what we classify as AI slop from just a few weeks ago," Srinivasan wrote.

For a coach who posts five times a week on LinkedIn, that is the whole ballgame. Reach is the funnel. And the most exposed group on the platform is the one that built a content habit on a prompt.

The number behind the button

LinkedIn didn't do this out of nowhere.

In July, AI-detection firm Pangram published a study that landed like a brick. The company scanned over a million social media posts and found AI content on every platform it checked. LinkedIn had the highest proportion: more than 40% of posts longer than 250 words were flagged as fully written by AI.

The distribution is what stings. Posts from LinkedIn made up roughly one-third of all content Pangram scanned, but accounted for 62% of all AI-flagged material in the entire dataset. Reddit, by contrast, was 36.7% of the scanned content and just 4.4% of the AI flags.

Pangram's own read: LinkedIn's built-in AI writing features likely made it worse.

Long-form thought leadership is the native format of the coaching industry. The 300-word origin story. The client breakthrough. The "three lessons from a hard quarter" post. That is precisely the content bracket Pangram measured.

What LinkedIn is actually suppressing

Here is the part most coaches are getting wrong, and it matters for what you do on Monday.

The target is not AI. The target is emptiness.

LinkedIn Creator Product Lead Sam Corrao Clannon: "We internally define AI slop as content that is potentially sophisticated or polished in its presentation, but lacks substance. So it doesn't have any particular experience, perspective, or insight, but is sort of just empty text that's posted to take up space and garner attention without effort on the other side."

He was blunt about the other half: "It's not anything that uses AI in any capacity. I want to be clear about that. There's a lot of people who use AI to refine their own thinking and insights...All that sort of stuff? Totally fine."

He also described the mechanism in a way that should change how you think about a single flag. "The feature again, it's not a report; it's a disinterest signal at the viewer level," he said. "We then use those to train our classifiers, but the classifier is really looking at things that have text without substance."

Corrao Clannon added that the option is not "a reporting path for policy violations," and that when a user flags something, it primarily affects that user's own feed. Srinivasan said no single piece of feedback determines how content is distributed.

So one annoyed reader doesn't bury you. A pattern does. The flags feed a classifier, and the classifier is now grading the entire feed.

Where to look to see if it happened to you

This is the detail that hasn't travelled far enough.

LinkedIn is adding a feature that notifies users when their own posts have been flagged as suspected AI spam. "This week, we're taking another step," Srinivasan said. "If a post receives enough community feedback, you may see a message in your Post Analytics to help you understand how your content is being received."

Not a warning email. Not a strike. A line inside your own analytics that only you see.

Which means thousands of coaches are currently being quietly throttled and have never opened the screen that would tell them. Go look. Post analytics, on your recent long-form posts, the ones you were proudest of and that mysteriously did 300 impressions.

LinkedIn killed its own writing button

The most revealing move wasn't the flag. It was what LinkedIn removed.

Alongside the slop-reporting option, the company said it was replacing its own AI writing feature with a proofreading tool. The "enhance your post" button that used AI to write for you is being swapped for one that proofreads your words instead of changing your voice — the platform effectively saying: use AI to fix typos, not to write your thoughts.

A platform retiring a feature it built and promoted is a platform that has read its own data.

LinkedIn also says it now blocks "hundreds of thousands" of automated comment attempts daily, plus millions of other automation attempts in recent months. The comment-pod economy that a lot of coaching businesses quietly rely on is standing in the blast radius.

The uncomfortable caveat

Detection is not a science.

Critics warn Pangram-style findings carry high false-positive rates, especially for text written by non-native English speakers. Researchers have consistently pointed out that AI detectors produce probabilistic estimates rather than definitive conclusions, and can generate false positives depending on writing style, topic and model.

And LinkedIn's flags aren't coming from a lab. They come from readers applying their own standards, which means running your own draft through AI is no guarantee they won't read it as machine-made.

You can write every word yourself and still get flagged for having a tidy structure and a three-word punchline.

Even the executive running the crackdown has caught himself. "I know I'm increasingly conscious on how to not sound like AI," Srinivasan wrote.

What the smart operators are doing

Forbes contributor Jodie Cook, who covers AI and founder-led growth, put the coaching-relevant version of it plainly. LinkedIn is hunting AI slop and most people feeding it don't know they've been caught, she wrote: the same three-word punchlines, the same voice that belongs to nobody, the post that bombs because the dream client scrolls past words that feel manufactured.

She isn't arguing for abandoning the tools. Cook says she grew her LinkedIn to 57,000 followers by experimenting with AI writing and studying the data, but that LinkedIn has changed the game and lazy AI posts no longer work.

The practical line is now drawn in a different place. The thinking has to be yours: the number from the client call, the thing you'd defend in a room of peers, the opinion that could cost you a follower. The polish can be outsourced.

"Despite the progress, we know we have more to do to ensure LinkedIn remains a place where you can find real people and real perspectives," Srinivasan wrote. "This all remains very top of mind."

Translation for anyone charging $5,000 a month to help other people find their voice: the feed is about to get very good at telling whether you have one.