Why LinkedIn is adding a button to flag AI content

LinkedIn has added a new option to its report menu that lets you flag a post as seeming like AI slop. When you click this, the site hides the post, and your feedback is sent to the company to help them identify similar content in the future. This is part of a larger cleanup effort at the platform. They are also updating their automated systems to block thousands of bot-generated comments and are introducing new classifiers—tools designed to automatically sort and filter out low-quality items—to keep your feed feeling human. Additionally, LinkedIn is tweaking its own writing assistant tools. Instead of using AI to rewrite your posts in a generic style, they are switching to a simpler proofreading tool that sticks to your original voice.

To understand why LinkedIn wants your help, you have to look at how these systems learn. Companies use a process called training where they feed an AI model—the digital engine that generates text—millions of examples so it can spot patterns. Currently, LinkedIn wants its systems to get better at spotting content that is mass-produced by AI. When you flag a post as slop, you are essentially providing a labeled signal. You are telling the system, this specific example is what we consider useless, repetitive, or inauthentic. By collecting thousands of these signals, LinkedIn can adjust, or tune, its internal models to recognize those specific patterns of robotic writing without needing to search for them manually. It is like teaching a new intern; at first, they might struggle to tell the difference between a thoughtful post and a generic one, but the more you correct them, the better they get at filtering for quality.

The rise of AI-generated content has created a crowded, noisy internet where it is increasingly hard to find genuine human connection. LinkedIn is betting that if they do not curb this trend, people will stop participating. By treating this as a top priority, they are signaling a shift away from just filling feeds with as much content as possible and toward filtering for what people actually find valuable. Ultimately, the future of these platforms might depend on their ability to prove that they are still spaces for real experts, not just high-speed automated output.