The option sits inside LinkedIn’s existing feed controls
The “seems like AI slop” option appears in the menu used to report a post or request less content like it. It gives low-quality AI content a named category rather than leaving users to fit it under broader feedback such as irrelevant or low-quality material.
Testing found that the specific option applies to posts but does not extend to comments. LinkedIn is also working on a private analytics view through which creators will be able to see when users have applied the label to their posts.
LinkedIn Chief Product Officer Hari Srinivasan described AI slop as a “top priority” and said people visit the platform to connect with “real people” sharing their own perspectives, ideas, and expertise. Giving the problem a dedicated label is an acknowledgement that generic AI output has become a distinct feed-quality issue for the professional network.
The flag changes your feed rather than removing a post
Flagging a post as AI slop suppresses it and similar posts from the reporting user’s feed instead of automatically taking the content down. The control therefore works primarily as a personal recommendation signal, although LinkedIn also intends to use the feedback to improve its ranking systems.
“Slop is hard to define and the definition changes; this lets us tune our models and make better feeds,” Srinivasan said. Translated out of product language: LinkedIn wants members to supply examples that its automated systems can learn to rank lower.
That approach accommodates the subjective nature of the label, but it also puts users in the quality-control loop. One person may object to any visibly automated prose, while another may reserve the term for repetitive, generic, or low-effort posts. LinkedIn’s system collects those individual judgements without treating each report as a moderation verdict.
LinkedIn is narrowing its own AI writing tool
LinkedIn says it is removing “enhance your post”, a feature that could generate or substantially rewrite text, and replacing it with a more limited proofreading tool. Srinivasan says the replacement “proofreads your words, but does not change your voice”.
The policy line LinkedIn is drawing allows AI to refine a person’s writing while discouraging AI from replacing the person behind it. Srinivasan said many users “refine thoughts” with AI and separated that practice from content that feels inauthentic, generic, or low-effort.
LinkedIn has not fully explained how the replacement will distinguish proofreading from rewriting in practice. Grammar, clarity, tone, and structure can overlap, leaving the eventual behaviour of the tool more consequential than its label.
Microsoft’s AI expansion leaves LinkedIn with the clean-up job
LinkedIn is confronting the feed-quality consequences of generative AI while its owner, Microsoft, continues investing billions of dollars in AI and integrating the technology across its products. The tension is structural: creation tools make polished text easier to produce, while a social platform depends on readers believing that posts still represent the people whose names appear above them.
LinkedIn previously offered a tool that could generate post text and is now asking members to identify output that feels automated or empty. Both positions can be rational. AI can help users express an original idea more clearly, while cheap mass production can overwhelm the useful material that gives a professional feed its value.
LinkedIn owns the feed, the ranking system, and the relationship with its members, so it also absorbs the cost when automated content degrades that experience. The new flag turns members into an additional source of ranking data, helping the platform separate acceptable assistance from content people actively want to avoid.
Users gain control, but creators face a subjective signal
LinkedIn users can use the flag to reduce similar posts in their own feeds, while creators may eventually see the resulting feedback privately in their analytics. The announced mechanism does not automatically remove flagged posts or establish a disclosed platform-wide penalty for using AI.
For creators, LinkedIn’s message focuses on authenticity rather than an outright ban on AI assistance. Original expertise that has been proofread with AI fits the distinction Srinivasan described; generic text that substitutes polish for a recognisable point of view is more exposed to negative feedback.
For readers, the immediate effect is modest but practical: fewer posts resembling the ones they flag. The larger test is whether those reports give LinkedIn enough reliable data to improve the feed without turning a subjective insult into a blunt ranking rule.
Leave a Reply