Over 54 percent of longer English-language posts on LinkedIn are likely AI-generated, according to data from a new analysis from AI detection tool Originality AI. Originally published in WIRED, this analysis showed a major increase in AI content on LinkedIn occurring from the start of 2023 onward. This timing aligns with the rise of ChatGPT, which was launched at the very end of 2022 and gained major traction in 2023.
Originality AI Detection Study
To conduct the analysis, Originality AI reviewed a sample of 8,795 public LinkedIn posts. In order to be included in the dataset, posts had to be over 100 words long and published between January 2018 to October 2024. Through the analysis, Originality AI found that a “negligible” number of posts appeared to be generated by artificial intelligence during the first years, through the end of 2022.
There was a huge spike in likely AI-generated posts at the beginning of 2023, with the number rising 189 percent. This number has since leveled off, and now roughly half of all longform English posts on LinkedIn appear to have been created with the help of artificial intelligence.
The Rise of AI Content on LinkedIn
LinkedIn, which is owned by Microsoft, is increasingly trying to present itself as an AI company. LinkedIn even offers LinkedIn Premium paid users access to its in-house AI writing tools that help them edit posts, messages, and profiles.
However, the company also says that it has “robust defenses in place to proactively identify low-quality, and exact or near-exact duplicate content,” according to Adam Walkiewicz, head of feed relevance. “When we detect such content, we take action to ensure it is not broadly promoted.”
WIRED interviewed multiple LinkedIn posters to find out why they turned to AI post generators. Some were non-native English speakers who used the AI writing tools to help fix grammar errors. Others turned to AI LinkedIn post generators to quickly create rough drafts that they then edited before posting to LinkedIn. Aspiring influencers also used post and comment generators to create a constant dream of daily content to feed the algorithm.
While some users disclose whenever they use AI to create or edit a post, others opted to keep that information confidential, making it difficult to tell whether any given post counts as one of the 54 percent of likely AI-generated posts.