LinkedIn Reports 40% Drop in Views for AI Content

LinkedIn claims to have reduced views on content labeled as "AI slop" by 40% and announces that over one million people have used its reporting button, while rolling out new detection tools and alert messages for authors.

LinkedIn announces a 40% decrease in views on content labeled "AI slop"

Hari Srinivasan, LinkedIn's product director, states that users are now seeing 40% fewer views on posts that the platform classifies as "AI slop" compared to a few weeks ago. LinkedIn has also introduced what it describes as "new and improved" classifiers to detect AI-generated content and has removed the feature that suggested "improving your post" with AI. Srinivasan emphasized that combating "AI slop" is a major priority for the company. Earlier this year, LinkedIn also announced its intention to crack down on comments produced at scale with little or no human intervention.

Reporting button surpasses one million clicks

According to Hari Srinivasan, over one million people have already used the "Seems like AI slop" button, accessible from the three-dot menu of a post. LinkedIn also plans to add a message that alerts authors when "certain members" believe a post appears to be AI-generated. Srinivasan notes that this feature was designed with the assumption of users' good faith and aims to provide useful feedback. He serves as the product director at LinkedIn.

Announcement of the button and context of a strong presence of generated texts

The "Seems like AI slop" button was announced on July 30. Prior to this announcement, the detection tool Pangram had determined that 41% of long posts on LinkedIn were flagged as entirely AI-generated, an analysis reported by 404 Media.