Study Reveals Alarming Levels of AI-Generated Content on LinkedIn and X
Key Takeaways
- ▸41% of LinkedIn longform posts are fully AI-generated; roughly 25% of X articles are fully AI-written, with another 23% AI-assisted
- ▸AI-generated content is particularly concentrated in longform posts across all platforms, reflecting AI's verbose nature and labor-savings benefits for longer pieces
- ▸Users encounter AI-generated content regularly on major platforms as part of their everyday browsing—not just on fringe or obscure sites
Summary
Browsing data from Pangram, an AI content detection company, reveals that AI-generated content has become pervasive on major social media platforms. According to analysis of roughly one million posts collected over two months using Pangram's Chrome extension, approximately 41% of longform content on LinkedIn and roughly one-third of longform posts on X are likely to be fully AI-generated, with an additional 23% of X content being AI-assisted. The study found that longer-form content is significantly more prone to AI generation across all platforms, including Reddit and Substack, where roughly 10% of longer posts are AI-generated.
Pangram's research challenges common assumptions about where AI-generated content appears, showing that 'AI slop' is not confined to obscure spam sites or SEO farms, but rather is actively encountered by users on the most popular social media platforms. The data reveals that professionals are more willing to use AI to compose posts on identity-linked platforms like LinkedIn compared to anonymous platforms. The study found that top-level posts are far more likely to be AI-generated than replies and comments underneath them, suggesting users are experiencing a persistent stream of algorithmically-generated content while browsing.
- Professionals use AI at higher rates in identity-linked professional settings, contradicting assumptions that AI adoption would be limited to anonymous platforms
Editorial Opinion
This data provides a sobering picture of how AI-generated content is contaminating mainstream social media. While AI tools promise productivity gains, the widespread adoption for professional posts suggests these platforms are becoming increasingly flooded with low-effort, algorithmically-optimized material. Unless platforms implement meaningful detection and moderation policies, users will continue facing an expanding 'tax on their attention' as they navigate through AI-generated content.
