- Study Finds Social Media Posts Are Increasingly Generated with AI
- LinkedIn particularly affected, with 40% of long posts written by AI
- Substack and Twitter/X were also severely affected
LinkedIn and other social networks are rapidly being consumed by spam posts written by AI, new research claims.
A report from AI discovery firm Pangram Labs found that nearly half of all long-form posts (250+ words) on LinkedIn were created entirely by AI, and companies like Substack and X/Twitter also saw a big increase in such content.
“LinkedIn was the most AI-saturated platform, with more than 40% of long-form posts being marked as fully AI-generated,” the company’s report said.
The study, which also examined Medium and Reddit along with other social networks for a data set of more than one million posts, found that one in four long-form posts on social media were marked as entirely AI-generated, and that longer-form content was much more likely to be created with AI than short-form content.
Pangram found that LinkedIn was the most AI-saturated platform, where more than 40% of long posts were flagged as fully AI-generated, with Substack being the least affected, and longer posts were often much less likely to be AI-generated.
LinkedIn was also identified as having the highest proportion of AI of all the platforms included in the report, as although its posts only accounted for a third of the items scanned, it accounted for almost two-thirds (62%) of all AI content flagged by the system.
However, when mixed AI and human content was included,
“Our data shows that AI-generated content is a problem across all platforms, and it’s hitting long-form content especially hard,” Pangram said.
“Contrary to what one might expect, people are overwhelmingly willing to use AI to speak on their behalf in professional settings associated with their real identity, and are less likely to use it on informal, anonymous platforms.”
“AI writing is now a problem everywhere in social media,” said Pangram Labs CEO and co-founder Max Spero. “An Internet that is completely inundated with undisclosed AI content is bleak, but we don’t think it’s inevitable.”
Follow TechRadar on Google News and add us as a preferred source to receive news, reviews and opinions from our experts in your feeds.
