LinkedIn Is Drowning in AI Slop — So They Built a Button For It

Open LinkedIn on any given day and you'll scroll past a familiar rhythm: a humble-brag hook, three short punchy lines, a bulleted list of "lessons," and a closing line begging for engagement. It reads suspiciously similar no matter who posted it, because increasingly, it wasn't really them — it was an AI writing tool doing the talking. The platform that used to be for job hunting and the occasional cringe update has become the most concentrated feed of AI-generated content on the internet.

The pattern is easy to spot once you know it: em dashes everywhere, a rhetorical question nobody asked, and advice so generic it could apply to literally any industry. It's not that AI-assisted writing is inherently bad — it's that so much of it is optimized purely for the algorithm rather than for saying anything real. That's the definition of slop we use on this site: content or products that exist to capture attention or a sale, with no actual substance behind them.

LinkedIn has apparently noticed too. The platform has started rolling out ways for users to flag posts as AI-generated or low-value, essentially crowdsourcing a slop filter directly into the feed. It's a tacit admission that the platform's own algorithm helped create the problem in the first place — reward engagement bait long enough, and you get a feed optimized for producing exactly that.

Whether a report button actually fixes anything is a different question. Flagging tools only work if enough people use them honestly, and if the platform actually weights the feedback into what gets shown. Rotten Tomatoes works because a panel of critics with reputations on the line render a verdict; a crowdsourced "this feels like AI" button on a feed optimized for engagement is a much blunter instrument. It might quiet the worst offenders. It's not going to fix the incentive structure that produced them.

It's also a preview of where this is headed everywhere, not just on LinkedIn. As more of what we read, buy, and evaluate gets touched by AI somewhere in the pipeline, "is this actually good or is this slop" stops being a niche question and starts being the default question for any product or piece of content. That's the whole premise of this site — B2B software, consumer apps, and yes, probably feed content eventually too. A button that lets you flag something is a start. A panel of people willing to actually look closely and say so in public is the part that's harder to fake.