LinkedIn is cracking down on AI slop and its algorithm can spot it 94% of the time

LinkedIn has had enough of the "it's not X, it's Y" posts. It is doing something about them.

The platform rolled out a major algorithm update that detects and suppresses AI-generated content lacking original perspective, claiming 94% accuracy in early tests.

Posts following predictable ChatGPT patterns, generic phrasing or obvious AI structure will quietly disappear from feeds beyond your immediate connections – still visible to people who follow you directly, but no longer spreading across the platform.

What gets flagged: engagement bait, recycled thought leadership with no original angle, and anything that follows a clearly formulaic AI construction. LinkedIn specifically called out the "it's not X, it's Y" post format as a target. If your content could have been written by literally anyone using a basic prompt, that is now a liability.

The important caveat: LinkedIn's 94% accuracy claim has not been accompanied by data on false positives. How often legitimate, original posts get wrongly demoted is not yet clear, and the rollout is gradual – the platform says it could take months before the changes are fully felt.

AI-assisted content is still fine, according to LinkedIn, as long as it contains original thinking and adds something to a conversation rather than just filling space.

The honest read on this: LinkedIn lost control of its feed to AI slop and is now trying to clean it up. For anyone who has been using the platform recently, that is overdue. For marketers and brands who relied on volume over substance, it is time to recalibrate. Reach now depends on having something worth saying.