
LinkedIn is the undisputed king of long-form AI slop, according to a study spanning five platforms
Quick Answer
A recent Pangram analysis reveals that 41% of long-form posts on LinkedIn are AI-generated, making it the leading platform for AI content.
Quick Take
A recent Pangram analysis reveals that 41% of long-form posts on LinkedIn are AI-generated, making it the leading platform for AI content. Despite only comprising a third of the posts analyzed, LinkedIn accounted for nearly two-thirds of detected AI content, while Substack had the lowest rate at around 10%. The study highlights the growing prevalence of AI in social media, prompting LinkedIn to tighten regulations on AI-generated posts.
Key Points
- 41% of long-form posts on LinkedIn are flagged as AI-generated.
- LinkedIn made up a third of posts but nearly two-thirds of AI content.
- X/Twitter had close to 50% AI-generated long-form articles.
- Substack had the lowest AI generation rate at around 10%.
- Pangram's detection model claims a false positive rate of only 0.01%.
📖 Reader Mode
~1 min readOne in four social media posts over 250 words is AI-generated, according to a Pangram analysis. LinkedIn tops the list with 41 percent of long-form posts flagged as AI-written. The platform made up only a third of all posts scanned but accounted for nearly two-thirds of all detected AI content. On X/Twitter, close to half of long-form articles were AI-generated or AI-assisted. Substack had the lowest long-form AI rate at around 10 percent. Reddit replies were 98 percent human-written, but standalone posts contained AI text far more often.

The data comes from Pangram's Chrome extension, which scanned over one million posts across five platforms between April and June 2026. The company claims its Pangram 3 detection model has a false positive rate of 0.01 percent, but it's likely better at identifying human-written content than AI-generated content, so the real AI rate could be even higher. The study makes no claims about content quality, but LinkedIn itself seems to feel the pressure and has already started cracking down on AI-generated posts.
— Originally published at the-decoder.com
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.

