LinkedIn just made “sounds fake” a first-class product signal. That matters for talent teams because your employer brand, recruiter outreach, and hiring-manager posts all live in the same feed where members can now privately flag content as AI slop.
Chief Product Officer Hari Srinivasan said the company is ramping the ability for members to report when a post or comment “seems like AI slop,” because slop is hard to define and the definition changes. Member reports help LinkedIn tune classifiers and clean suggested content from outside your network. Creators will also get a private analytics flag when members feel a post came off as inauthentic or heavy on AI.
“We want members to get feedback from real humans on what sounds authentic — not just have an AI detector review it and get it wrong.”
That is Hari Srinivasan, explaining why LinkedIn is betting on human flags instead of a silent AI-only detector.
What shipped, and what LinkedIn is killing
The practical product moves for TA to track:
Report path: three-dot menu on posts and comments includes a “seems like AI slop” style flag (rolling out in test).
Defense layer: LinkedIn says it is catching hundreds of thousands of automated comment attempts daily and has blocked billions of other automation attempts (posting at scale, slop) in recent months.
Classifiers: new models aim to cut AI-slop and low-quality content in suggested / out-of-network inventory.
Creator feedback: private analytics nudge when members rate your post as inauthentic or AI-heavy.
Writing tools pivot: LinkedIn is removing “enhance your post” (the rewrite-your-voice helper) and replacing it with proofreading that keeps your wording.
Srinivasan’s own diagnosis of why people leaned on AI for posts: LinkedIn is not a one-word kind of place, so people felt safer running drafts through a model. The product answer is not “ban AI.” It is “stop rewriting the human voice.”
Profile and page verification keeps expanding so members know who they are reading. That trust stack only works if the words still sound like a person.
