- LinkedIn members can flag posts and comments that “seem like AI slop,” supplying feedback the company says will help tune its feed models. [1] - Flagged posts are hidden for the reporting member, while classifiers can reduce low-quality content in recommendations and outside-network feeds. [1][2] - LinkedIn will replace its AI writing enhancer with proofreading and test private authenticity feedback for creators, but it has not disclosed report thresholds or appeal safeguards.
[1] LinkedIn is adding a “Seems like AI slop” option to posts and comments, turning members into another source of training signals for systems designed to suppress generic, automated content. Chief Product Officer Hari Srinivasan announced the changes on July 30, saying LinkedIn is expanding classifiers that identify AI slop or otherwise low-quality posts.[1]
The option appears in a post’s three-dot menu. Selecting it hides the post for that member and produces a message thanking them for the feedback, according to The Verge’s test. LinkedIn said the reports will help it tune its models and reduce unwanted material in suggested posts and content from outside a member’s network.[1][2]
Downranking rather than a blanket AI ban #
LinkedIn defines AI slop as low-effort, AI-generated material that may sound polished but lacks an original perspective or substance. Its systems also target automated comments posted at scale and responses that merely restate a post. Content judged to be AI-generated and lacking perspective can remain available to direct connections while receiving less distribution elsewhere.[3]
That standard does not prohibit all AI assistance. Srinivasan said AI use and slop are different, and LinkedIn will remove its “enhance your post” feature from post and message composers. A replacement will proofread members’ words without rewriting their voice.[1]
Safeguards and rollout details remain limited #
LinkedIn plans to test private notices in creators’ analytics dashboards when members consider a post inauthentic or heavily AI-written. The company has not said whether creators will see report totals, what threshold triggers a notice, or whether reports directly affect distribution.[1]
As of July 31, LinkedIn had not identified rollout markets, a completion date, protections against coordinated false reporting, or an appeal process for this feedback. The reports appear to function initially as ranking data rather than formal findings that a post violated platform rules.[1][2]
The change follows a Pangram analysis of more than one million social posts scanned through an opt-in browser extension. The company said more than 40% of sampled long-form LinkedIn posts were classified as fully AI-generated, though that user-selected sample is not a platform-wide audit.[4]
Companies mentioned #
Further sources #
[1] Hari Srinivasan, “AI slop is a top priority for all of us,” LinkedIn, July 30, … ↗
[[2] Jay Peters, “LinkedIn actually adds a ‘seems like AI slop’ button,” The Verge, … ↗](https://www.theverge.com/ai-artificial-intelligence/973384/linkedin-seems-like-ai-slop-button)
[[3] LinkedIn Corporate Communications, “Keeping conversations real on LinkedIn,” Ju… ↗](https://news.linkedin.com/2026/keeping-conversations-real-on-linkedin)
[[4] Max Spero and Annamieka Aerts, “AI Content Is Everywhere on Social Media, Espec… ↗](https://www.pangram.com/blog/ai-in-your-feed)
The stories that matter, in one email. Free — unsubscribe anytime.