LinkedIn Adds 'Seems Like AI Slop' Button to Flag Posts

LinkedIn now lets you tell it, in so many words, that a post looks like it was written by a chatbot. The platform has added a 'seems like AI slop' option to the three-dot menu on posts, and selecting it hides the post from your feed. 404 Media confirmed the button is live through its own tests.

The company's framing, from chief product officer Hari Srinivasan, is that user reports feed a detection pipeline. 'AI slop is a top priority for all of us,' he said, adding that people come to LinkedIn 'to connect with real people and share their real perspectives, ideas and expertise.' The platform says it is 'ramping up a series of new and improved classifiers that identify if a post is AI-slop or generally low-quality content,' with the button's clicks acting as a training signal.

The awkward context is that LinkedIn is also, itself, in the AI writing business. The same product offers AI-powered writing assistance to users, an irony 404 Media notes directly. And the scale of the problem, if you take one recent third-party estimate at face value, is not small: the AI detection service Pangram estimated that 41 percent of long-form posts and 30 percent of short-form posts on the platform are likely AI-generated.

The honest caveat is that 'slop is hard to define and the definition changes,' as LinkedIn itself acknowledges. Any classifier trained on user reports will inherit whatever biases the reporters bring, and the reporting does not tell you how aggressive the hiding actually is, whether flagged authors get penalized elsewhere in the feed, or what the false-positive picture looks like, particularly for non-native English writers.

For anyone whose distribution depends on LinkedIn, the interesting question is how a real detection push reshapes what actually reaches feeds. If the classifiers work, the platform starts to feel less generic. If they overshoot, a lot of legitimate corporate prose starts disappearing into a quiet penalty box.