We have likely all experienced the same feeling on LinkedIn within the last couple of years… seeing a long-form post and suspecting it of being AI-generated but the author is passing it off as their own thought leadership.
In this study, we look at the impact of ChatGPT and other generative AI tools on the volume of AI content that is being published on LinkedIn.
Our process was to analyze 8,795 LinkedIn long-form posts using our AI detector to determine if the post was likely AI-generated or likely human-written. The study spanned content published over an 82-month period from January 2018 to October 2024, which we analyzed to identify trends in AI usage. These long-form posts contain at least 100 words and were curated from different LinkedIn users on the platform.
Our AI detector is the most accurate AI detector on the market and has been verified by both our own study and 8 published 3rd party studies.
When we say "likely AI-generated" here is what our AI detector has been trained to call AI generated vs human generated:
*AI Outline is defined as using AI (an LLM) to create a content idea, do some research, and/or create an outline. The level at which AI is used during this process may vary and could potentially affect the likelihood the text is detected as AI or human
ChatGPT’s release may have influenced a surge in AI-driven content on LinkedIn. Our result shows that prior to the release of ChatGPT in 2022, the use of AI by LinkedIn users was at a low.
After the release of ChatGPT, the amount of AI content on LinkedIn posts per month significantly spiked, showing an 189% increase from January to February 2023.
This spike is likely due to ChatGPT’s ease of use, accessibility, and ability to generate high-quality human-like content.
After this significant surge in February 2023, there has been a stabilization that might indicate a new normal for LinkedIn’s content composition, where roughly half the content could be AI-generated. The release of ChatGPT marks a dramatic turning point — this spike represents a notable shift in content, where AI became a mainstream topic, likely due to ChatGPT's impact on the public's awareness of AI.
Looking at LinkedIn posts published recently, we can see that over 50% of all long-form LinkedIn posts are suspected of being AI-generated.
We analyzed LinkedIn posts published in October 2024 from our sample data and learned that a significant portion of LinkedIn's long-form content creation is AI-assisted.
This was also the trend between June and October 2024, which shows a new norm on the platform.
Since the release of ChatGPT, which likely caused a surge in AI content on LinkedIn, the length of LinkedIn posts has significantly increased. On average, LinkedIn posts have a higher word count.
The chart above shows a correlation between LinkedIn long-form post word count and suspected AI presence over time. The trend here likely confirms the use of AI tools not only increased the percentage of AI-generated posts per month but also influenced the word count of these posts. 2023 and 2024 show a strong correlation between AI usage and word count in LinkedIn long-form posts. This highlights how AI is shaping the dynamics of LinkedIn content creation.
Here are some specific examples of people admitting to using AI in their LinkedIn posts…
While AI is clearly being used, it is often not admitted to being used by users. The difficulty of generating new content consistently is highlighted in the article - “I Used AI to Publish on LinkedIn for 7 Days – Here's What Happened” which provides clear evidence of AI being leveraged.
Another example of AI usage in LinkedIn is its application in various ways. For instance, the YouTube video “Build a LinkedIn Marketing Strategy with ChatGPT” demonstrates how LinkedIn users can create LinkedIn posts that try to genuinely reflect their expertise while leveraging ChatGPT.
For this segment of the study, we looked at 2,726 posts from the dataset that were over 100 words long and were published after the launch of ChatGPT (from Dec 2022 to Oct 2024).
Posts were classified as Likely AI or Likely Original using Originality.ai’s AI detector.
Comparing the engagement of each post (likes and comments each count as one engagement), we see that the average likely-AI-generated post received 45% less engagement than a Likely Original post.
The explosive growth of AI posts on LinkedIn has long been suspected and the implications for it as the platform for thought leadership is being challenged when over ½ of the posts on it are likely AI-generated.
Not sure whether a post you’re reading is AI or human-written? Try the Originality.ai AI Detector today.
We believe that it is crucial for AI content detectors reported accuracy to be open, transparent, and accountable. The reality is, each person seeking AI-detection services deserves to know which detector is the most accurate for their specific use case.
With our proprietary Originality.ai AI detection tool, we analyzed the presence of AI in health and wellness reviews across key consumer industries — baby formula, skincare, and health supplements. These are our findings.
