Study: 41 Percent of Long LinkedIn Posts Are AI-Written — the Highest of Any Platform
AI-detection company Pangram scanned over a million social media posts. One in four long posts across platforms is AI-generated — on LinkedIn it's four in ten. Here's how to read those numbers.
If your LinkedIn feed has been feeling a bit samey lately, there’s now data behind the hunch. According to an analysis by Pangram, a company that builds AI-text detectors, 41 percent of LinkedIn posts longer than 250 words are AI-generated — the highest share of any platform studied. Across all five platforms in the study, roughly one in four long-form posts was flagged as AI-written.
The numbers come from Pangram’s Chrome extension, which scanned over one million posts between April and June 2026. Some highlights: LinkedIn made up only about a third of all posts scanned but accounted for nearly two-thirds of detected AI content. On X (formerly Twitter), close to half of long-form articles were AI-generated or AI-assisted. Substack, the newsletter platform, had the lowest long-form rate at around 10 percent. And Reddit showed a curious split — replies were 98 percent human, while standalone posts contained AI text far more often.
What’s behind this? Two things worth understanding. First, incentives: LinkedIn rewards frequent posting with professional visibility, and AI makes frequent posting nearly free — so the platform gets flooded with what’s now commonly called “slop,” low-effort AI content produced at scale. LinkedIn itself has reportedly started cracking down on it. Second, a caveat about the measurement: AI detectors are never perfect. Pangram claims a false-positive rate of 0.01 percent — the share of human posts wrongly flagged as AI — but detectors are generally better at confirming human writing than catching all AI writing, so the true AI share could be even higher than reported. Also worth noting: the study says nothing about quality. An AI-assisted post isn’t automatically bad, and a human post isn’t automatically good.
What this means for you: As a reader, a bit of healthy skepticism is now realistic, not cynical — especially for polished, list-heavy posts from accounts that publish suspiciously often. As a writer, the bar for standing out has weirdly dropped: specific experiences, real numbers, and a recognizable voice are exactly what the slop wave can’t fake. And if you use AI to help you write — plenty of people do — the study is a reminder that readers are increasingly primed to notice when there’s no human substance underneath.
Sources
Google's SensorFM Wants to Be One AI Brain for All Your Wearable's Health Data
Google Research trained a foundation model on over a trillion minutes of Fitbit and Pixel Watch data from five million people. It beats specialized models on 34 of 35 health tasks — and hints at where AI health assistants are heading.
