LinkedIn’s new “Seems like AI slop” button has been clicked by more than one million people, according to the company. The feature, which appears under the three-dot menu on each post, is a trust tool designed to let members flag content that looks as if it was written by artificial intelligence with little or no real expertise behind it. LinkedIn shared the milestone in a post from chief product officer Hari Srinivasan, who also said the change is already having an impact on feed quality.

Why LinkedIn introduced a “Seems like AI slop” button

LinkedIn announced the button on July 30. It was part of a response to the amount of generic AI-written material that had started to fill professional feeds. According to an analysis by the AI-detection company Pangram, 41 percent of long-form LinkedIn posts were being identified as fully AI-generated. That statistic drew attention to a growing problem: as language models became better at producing plausible business content, LinkedIn’s feed increasingly became home to formulaic posts about leadership, productivity, personal branding, and other broad professional themes.

The button asks users to make a subjective call. Instead of saying “This violates our rules,” it simply says “Seems like AI slop.” That wording matters. It allows LinkedIn to gather signals from its community while also acknowledging that a post may only look automated to one reader and authentic to another.

The announcement also included an admission that LinkedIn’s own AI tools may have contributed to the problem. The company removed a feature that would “enhance your post” with AI. That tool had been designed to help members polish their writing, but it could produce the kind of buzzword-heavy prose that many users now identify as AI slop. By removing it, LinkedIn acknowledged that AI-assisted writing can easily create posts that sound more like a machine than a person.

What LinkedIn says the new data shows

Srinivasan’s Thursday post revealed that “over a million people” have clicked the AI slop button. He also said users are now experiencing “40% less views on what we classify as AI slop from just a few weeks ago.” LinkedIn did not explain exactly how the reduction was calculated, but it appears that suspected AI-generated posts are being distributed less frequently in feeds. The company seems to be using a combination of community feedback and automated classifiers to decide what to suppress.

LinkedIn also introduced “new and improved” classifiers to identify posts as AI. A classifier may look for patterns such as unnatural sentence rhythm, excessive listicles, vague references to “my journey,” or an overabundance of clichés. Those signals can be combined with user flags to decide whether a post should receive reduced reach.

The 40 percent reduction in views suggests that LinkedIn is not simply deleting content. Rather, it is downgrading it. That is a common content moderation approach used by many platforms when they do not want to make a direct judgment about whether a post was made by AI, but still want to reduce its visibility.

Authors will be told if their posts seem generated

LinkedIn is also adding a message for people who publish posts that are flagged. The message says: “Some members told us this post seems like AI.” This appears to be an educational warning rather than a ban. Srinivasan said LinkedIn approached the feature “assuming good intent.” He added: “I know I’m increasingly conscious on how to not sound like AI & the goal is to provide helpful feedback.”

The warning does not accuse someone of violating a rule. It simply informs them that their audience is interpreting their writing as artificial. In some cases, this may encourage users to add more specific, personal detail to their posts. In other cases, it may push them to disclose when they have used AI to draft or edit text.

There is also an earlier part of LinkedIn’s moderation effort. Earlier this year, the company said it would crack down on comments users create at scale “with little or no human involvement.” Those comments, often called automated responses, are a form of AI slop because they add no information to a conversation. The new button is part of that broader push to reduce synthetic engagement and keep discussions meaningful.

A professional network cannot ignore AI slop

LinkedIn differs from other social media because its users expect a certain level of professionalism. A feed filled with fake success stories, hollow congratulations, and AI-generated career advice is not just annoying; it can make the platform feel less trustworthy. If users believe that advice on LinkedIn is written by bots, they may stop reading or stop sharing their own experiences. That threat is especially serious for a company whose core product is trust in professional relationships.

AI slop tends to be designed for engagement. Generic posts with broad phrases such as “Unpopular opinion” or “What I learned from failure” can attract comments because they are easy to respond to. AI makes it cheap to create these posts at scale, and some marketers and personal branding coaches have embraced generative tools. This creates an environment where human users are competing with a relentless supply of automated content.

By letting humans click “Seems like AI slop,” LinkedIn is making its members part of the detection system. Community signals have long been used by social platforms to identify spam, misinformation, and harassment. In this case, the signal is being used to tune feed ranking rather than to delete content. This lowers the risk of being accused of censorship while still reducing the reach of low-quality posts.

Where the AI slop button could lead

The success of the feature will depend on whether it can distinguish between AI-generated slop and legitimate writing from humans. Many job seekers, students, non-native English speakers, and busy executives write in direct, slightly formal language. Some of them may be falsely flagged because their writing style is clear and structured. LinkedIn’s classifiers will need to be trained on a wide variety of human voices to avoid punishing people simply because they do not sound conversational.

There is also a question about whether one million clicks means the tool is helping. One million clicks is a large number, but LinkedIn has more than one billion users. A million clicks could represent a small but engaged group of moderators. LinkedIn has not said how many posts were flagged, how many flags were needed before a post was labeled, or how many authors received the new AI warning. Without those details, it is difficult to know whether the signal is accurate.

Still, LinkedIn’s decision is a notable moment in the larger debate about AI content. Other platforms are also trying to deal with synthetic posts, but few have introduced an explicit “seems like AI” reporting option in such a public way. The button gives users a way to influence the algorithm with small actions. It also sends a signal to people who generate content for LinkedIn: if your posts feel robotic, the community can push them down and will tell you.

For now, LinkedIn’s approach is to reduce reach rather than remove posts. By doing this, it can preserve free expression, avoid accusations of banning human authors, and still make the feed more useful for people who want to see work, lessons, and stories shared by real professionals.

Source: The Verge News