LinkedIn introduces a button to address its biggest problem
LinkedIn is tackling the epidemic of AI-generated junk content by introducing a new flagging feature. The company is also restricting its internal AI post-generation tools to prioritize authentic user engagement.
If you have also noticed that your LinkedIn feed has been filled over the past year with generic posts about "why traffic jams on the Ayalon Highway are actually the best workshop for solving product bottlenecks" or "I lost an AirPods earbud. Here is what it taught me about the importance of redundancy"… you are definitely not alone. The professional social network is suffering from an epidemic of AI-based junk content, or in short, "slop," and the management there has had enough. Now they have a plan to bring real users and real content back to the center.
Start flagging
A study by the company Originality.ai, published last month, revealed an unsurprising figure: 81% of long posts on LinkedIn (>100 words) were created using AI — an increase from 50% in 2025. In response to this flood of content, LinkedIn is now offering a way to reduce AI junk and is launching a new button in the posts' settings menu ("hamburger").
The new button will allow you to flag posts that "look like AI slop," with the goal of helping the company refine its filters and reduce the amount of generic content in the feed.
"AI slop is an area of paramount importance for all of us," noted Hari Srinivasan, Head of Product at LinkedIn, in a post he published on the platform. According to him, "users come to LinkedIn to connect with real people and share their real insights, their ideas, and their expertise."
Source: LinkedIn
Alongside the button, LinkedIn has decided to cancel a previous feature that allowed for generating posts directly within the platform. Instead, the use of AI will be limited to proofreading only — without adding superlatives or unnecessary insights.
Clicking the button will remove the post from the feed and transfer it to a database that will train the filtering filter. Additionally, Srinivasan noted that the company is working on a mechanism that will notify users that their posts have been flagged as non-authentic, in the easiest and least offensive way possible.
