A recent study conducted by Semrush, analyzing 89,000 LinkedIn URLs referenced by AI tools such as ChatGPT, Google AI Mode, and Perplexity, has unveiled key factors that enhance visibility in AI-generated responses.
Scientists confirm: This is the most effective way to get your cat’s attention, according to new research
Elderly Couple Refuses Reserved Seats—Viral Train Standoff Sparks Fiery Debate on Courtesy
Summary
AI search engines represent a new frontier for visibility for brands and professionals: being mentioned in a response by ChatGPT or Perplexity means being recognized by an increasing portion of users. LinkedIn is among the most frequently sourced platforms by these tools. But what kind of content actually promotes these mentions? Semrush has explored this question and published a study based on 89,000 LinkedIn URLs from 325,000 prompts submitted to ChatGPT Search, Google AI Mode, and Perplexity between January and February 2026, in collaboration with LinkedIn.
LinkedIn: A Preferred Source for Generative AI
With an average citation rate of 11% across the three platforms studied, LinkedIn ranks as the second most referenced domain in AI-generated responses, trailing behind Wikipedia and YouTube. The variance between models is notable: Perplexity references LinkedIn in 5.3% of its responses, compared to 13.5% for ChatGPT Search and 14.3% for Google AI Mode.
Beyond the frequency of citation, the study also measures the semantic closeness between the AI-generated answers and the LinkedIn content referenced. Semantic similarity scores range from 0.57 to 0.60 depending on the model, indicating that AIs do not merely link to LinkedIn content; they actively capture its essence. For comparison, these scores are higher than those observed for Reddit (0.53-0.54) or Quora (0.435) in previous Semrush studies. In essence, this means that content published on LinkedIn can directly influence how a brand or topic is described in AI responses.
LinkedIn: What Content Types Are Most Frequently Referenced by AI?
Long-form articles are the predominant format in AI citations: they account for between 50% and 66% of LinkedIn URLs referenced depending on the model. Short posts come in second (15% to 28%), and company pages are also mentioned, albeit in varying proportions.
In terms of length, articles ranging from 500 to 2,000 words make up 72% to 77% of the citations, depending on the model. For posts, the 50-299 word range is most common (71% to 75%). A notable difference sets Perplexity apart from the other two models: it references company pages more frequently (59% of its LinkedIn citations), while ChatGPT Search and Google AI Mode favor content from individual members (59% in both cases). This divergence suggests that an effective LinkedIn strategy for AI visibility should involve both corporate page publishing and individual posting.
Relevance and Consistency Over Fame
The study identifies two author signals particularly correlated with AI citations. The first is publication frequency: 71% to 77% of cited post authors publish regularly (more than 5 posts in the preceding four weeks). The second is having an established community: nearly half of the cited authors have more than 2,000 followers. Notably, creators with fewer than 500 followers are cited at a comparable frequency to those with more, indicating that content credibility takes precedence over audience size.
The editorial intent of cited content confirms this relevance logic: between 54% and 65% of referenced posts aim to share knowledge or practical advice. Promotional content does exist in citations but remains in the minority, with a share ranging from 14% to 25% depending on the model. The originality of the content plays a crucial role: about 95% of cited content are original publications, with reshares only making up 5% of citations.
Engagement: Helpful, But Not a Deciding Factor
The engagement data of cited posts are clear: the median number of reactions lies between 15 and 25 depending on the platform analyzed, and the median number of comments ranges from 0 to 1. Posts that garner thousands of likes are not cited more frequently than others. AIs select the most relevant content in relation to the query posed, not the most popular.
This finding aligns with the conclusions of an Ahrefs study on sources cited by AI: AI search engines do not replicate the popularity logic of traditional search engines.
Similar Posts
- Study Reveals: 86% of AI Quotes Depend on Brand-Controlled Channels
- LinkedIn: The Second Most Cited Source by AI Technologies!
- Adobe Unveils LLM Optimizer: Boost Your ChatGPT SEO Rankings Now!
- Top 50 Websites Dominating Traffic With AI: See Who’s Leading the Digital Revolution!
- Why LinkedIn Posts Are Getting Fewer Views: Insights Into Algorithm Changes

Jordan Park writes in-depth reviews and editorial opinion pieces for Touch Reviews. With a background in UI/UX design, Jordan offers a unique perspective on device usability and user experience across smartphones, tablets, and mobile software.