If your LinkedIn routine was three rough lines and a tap on Enhance, that routine is over.

LinkedIn has retired "Enhance your post," the AI feature that could rewrite an entire draft for you. It was pulled in August 2026, right as the platform's AI slop crackdown was ramping up.

In its place, Premium members now get Post Proofreader. It reads the draft you wrote and suggests fixes to language and clarity, with each change shown in inline tracking so you accept or skip it one by one. It will not invent new ideas or regenerate the post.

That last sentence is the whole story.

What actually changed

LinkedIn's own description is deliberately modest. "Post Proofreader helps eligible Premium members improve post drafts before publishing. You can review suggested changes and decide whether to apply them."

LinkedIn added: "Suggestions appear with inline change tracking, so you can review what changed and stay in control of your content."

The suggestions look like tracked changes in Microsoft Word, and Premium members get three upgrade options when enhancements are recommended. It is the reverse of the old feature: previously a vague prompt could produce an entire post, whereas the proofreader only sharpens language you wrote yourself.

Social Media Today read the swap bluntly — rather than offering a way to re-create posts with AI, which made it easier for users to misrepresent themselves, LinkedIn is rebranding the process as proofreading, focused on improving language rather than generating a thought that didn't originate with the user.

Why it isn't a standalone feature update

This landed inside a larger enforcement push, not next to it.

On July 30, LinkedIn added a "Seems like AI slop" option to the post menu across the feed. Those flags feed LinkedIn's classifiers, and flagged content can lose distribution beyond your immediate network. Reporting has since been expanded to comments.

The early numbers are the part worth sitting with. As of September 11, 2026, a LinkedIn employee reported more than a million clicks on the slop button since launch and around 40% fewer views on content LinkedIn classifies as AI slop — and posters may now see a feedback note in their own post analytics.

So one hand builds a crowd-sourced detector that costs you reach. The other hand removes the tool most likely to trigger it. Read together, it's a position, not a product roadmap.

The honest version for coaches

A lot of coaching "thought leadership" on LinkedIn has been running on a generation step. Type a fragment, press a button, publish something that sounds like everybody else's Tuesday.

The input requirement just changed. You now have to bring a draft. Not a polished one — the proofreader exists precisely to clean up a rough one — but an actual draft with an actual opinion inside it.

That has a real cost, and pretending otherwise would be dishonest. If you were posting five times a week on generated text, five times a week is probably not survivable at the same effort level. Three posts you actually wrote will do more for you than five that a classifier is quietly throttling.

The secondary risk is the ghostwriting one. Executive ghostwriting is the highest-trust, highest-risk product most comms teams sell, and a CEO post that gets crowd-flagged is not a reach problem you fix next quarter. Coaches who write LinkedIn content for clients as part of a retainer are now exposed to the same thing, on someone else's name.

Nothing stops you using ChatGPT in a separate tab. LinkedIn isn't policing your drafting process; it's policing the output and the reach of it. But the platform has now removed the frictionless version from inside its own editor — and attached a visible cost to publishing text that reads like nobody was home when it was written.

The button is gone. The blank box is back.