Nearly 40% of long-form posts published on LinkedIn are fully generated by artificial intelligence, according to new research by AI detection startup Pangram, prompting the Microsoft-owned platform to introduce new measures aimed at reducing low-quality AI content.
LinkedIn has rolled out a new feature that allows users to flag posts they believe were primarily created by AI. The "Seems like AI slop" option appears in the three-dot menu on every post and feeds into the platform's content moderation and recommendation systems.
Chief Product Officer Hari Srinivasan said the feedback will help LinkedIn refine its detection models and improve users' content feeds.
Flagged posts may see reduced distribution, similar to content marked as "Not Interested," while creators receive a private notification in their analytics dashboard indicating that their post may have been perceived as inauthentic.
Platform expands AI moderation
According to LinkedIn, the company blocks hundreds of thousands of automated comment attempts every day, while millions of automated actions have been prevented over the past few months.
The platform is also discontinuing its AI-powered "Enhance Your Post" writing assistant and replacing it with a proofreading tool designed to preserve a user's original writing style.
Broader industry response
The rise in AI-generated content is prompting wider action across the technology industry.
Newsletter platform Substack recently introduced an AI detection feature through a partnership with Pangram, which recently secured $9 million in funding to expand its detection technology.
Meanwhile, Facebook is expanding its profile and page verification tools while introducing new options that allow users to block unwanted comments from company pages.
The issue extends beyond social media. Cloudflare has reported that automated bot traffic now exceeds human internet traffic, highlighting the growing challenge of distinguishing authentic online activity from AI-generated content.
