Over 40% of LinkedIn Long-Form Posts Identified as AI-Generated
According to a survey conducted by Pangram across five social media platforms, approximately one in four long-form posts were identified as AI-generated. LinkedIn showed the highest rate among surveyed platforms, with 41% of long-form posts identified as AI-generated. Given that the detection model employs conservative criteria, the actual proportion may be significantly higher.
A survey has revealed that one in four long-form posts on social media are being fully generated by artificial intelligence. The survey was conducted by Pangram, and analysis of data across multiple platforms showed that LinkedIn demonstrated particularly high proportions.
On LinkedIn, 41% of long-form posts were identified as AI-generated. Although LinkedIn accounts for approximately one-third of all scanned posts, it comprises roughly two-thirds of all detected AI content. This significant disparity between post volume and detection volume indicates that AI-generated content is disproportionately concentrated on LinkedIn compared to other platforms.
Pangram explained that it used an AI detection model designed to minimize false positives in its analysis. Because detection criteria were set conservatively, the actual rate of AI generation may exceed this figure. The 41% figure is considered a conservative estimate approaching the lower bound.
As a business-focused social media platform, LinkedIn hosts many long-form posts about careers and professional expertise. Such texts demanding a "professional tone" naturally overlap with formats at which generative AI excels, creating an environment where users are incentivized to leverage AI actively. The platform's emphasis on professional communication may ironically be driving the high rate of AI-generated content.
The increase in AI-generated content on social media raises fundamental questions for users about information credibility and author intent. In business contexts particularly, when LinkedIn posts viewed by recruiters and business partners are heavily reliant on AI generation, the extent to which they reflect the poster's actual capabilities becomes ambiguous. This issue extends beyond LinkedIn, representing a common challenge across all social media platforms where generative AI has proliferated.
Going forward, attention will focus on how platforms address AI-generated content. Currently, company policies vary, and clear standards for labeling requirements and detection tool implementation remain undefined. External research by organizations like Pangram continues to serve the visibility function, and industry-wide standard-setting is emerging as the key focal point.
This article is an original work independently written and edited by the AI issue editorial team based on factual reporting. © AI issue. Unauthorized reproduction, redistribution, or use for AI training is prohibited.