LinkedIn has become the most AI saturated major social media platform, according to new research from AI detection company Pangram. The study found that more than 40 percent of LinkedIn posts longer than 250 words were flagged as being fully written by artificial intelligence, the highest rate recorded among the platforms examined.
Pangram gathered the underlying data by asking users of its Chrome extension to opt in to sharing their browsing activity with the platform. Using its AI detection model, which the company says carries a 0.01 percent false positive rate, Pangram scanned nearly one million posts that users scrolled past across the internet over a two month period. Crucially, the study only scanned posts users actually encountered in their feeds, meaning the findings reflect AI content on the mainstream platforms people frequent daily, including LinkedIn, X, Medium, Reddit, and Substack, rather than being limited to spam sites.
While LinkedIn accounted for roughly a third of all the content Pangram scanned, posts from the platform made up nearly two thirds of everything flagged as AI generated, underscoring just how disproportionately concentrated AI writing has become on the professional networking platform. Pangram also found that top level LinkedIn posts were 1.35 times more likely to be AI generated than a comment on the same platform, though LinkedIn comments were still slightly more likely to be AI generated compared to top level posts on other platforms.
The pattern extended beyond LinkedIn as well. On X, a quarter of long form content was found to be fully AI written, with a further 23 percent showing signs of AI assistance. Substack recorded the lowest rate of AI generated content among the long form platforms studied, yet still saw more than a fifth of its posts flagged as either AI generated or AI assisted, suggesting the presence of AI writing has become widespread across most major platforms rather than concentrated in any single corner of the internet.
LinkedIn already includes a built in Enhance Post feature that offers users AI assistance when writing, making it simple for members to incorporate AI generated language into their posts. Responding to the findings, LinkedIn pointed to a statement from its global editorial executive editor Laura Lorenzetti, who said the platform has observed a rise in what many describe as AI slop, low effort AI generated content that can sound polished on the surface but lacks genuine perspective or substance. Lorenzetti said that when AI is overused, particularly at scale and in an automated manner, it dilutes the value of authentic human conversation on the platform, and confirmed that LinkedIn is working to detect and downrank content that appears AI generated so that members see less of it in their feeds going forward.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.