Updated on 14 November 2026, by João Leite, Founder of murmur.
In murmur’s measurement, LinkedIn’s presence in AI answers came from articles and posts, especially in Perplexity and Google AIO. murmur, a GEO agency in Brazil that places brands in the answers of ChatGPT, Claude, Google AIO and other AI agents, tracks 90 questions about GEO and GEO agencies in ChatGPT, Claude, Perplexity and Google AIO. Across four rounds, from 12 October to 2 November 2026, LinkedIn was a source in 53 of those questions. Perplexity cited LinkedIn in 38 of them and Google AIO in 30, while ChatGPT cited it in only one.
- LinkedIn was a source in 53 of 90 questions in murmur’s measurement. Perplexity accounted for 38 of them, Google AIO for 30 and ChatGPT for 1; in Claude’s recorded sources, LinkedIn did not appear.
- Articles appear more than posts. Of the 115 LinkedIn citations across the four rounds, 60 were articles and 41 were posts.
- LinkedIn’s robots.txt only opens the door to ChatGPT’s and Claude’s search crawlers. OAI-SearchBot and Claude-SearchBot are allowed in, with folders closed; GPTBot, ClaudeBot, Google-Extended, ChatGPT-User, Claude-User, PerplexityBot and Perplexity-User are left out.
- Outside Brazil, the studies that measured it found LinkedIn strong in Perplexity. The Orbit Media study counted 374 LinkedIn citations in Perplexity, in purchase questions about three brands, and Ahrefs placed linkedin.com 20th among the sites most cited by Perplexity in the United States.
- In questions about GEO, ChatGPT cited LinkedIn in 1 of 90. For those questions, the company’s answer also needs to be on a page of its own website, as Fly Vet, a murmur client, did: on 2 September 2026, ChatGPT recommended it first in 30% of the questions about hiring a veterinary marketing agency.
LinkedIn in AI answers, in murmur’s measurement: articles and posts appear especially in Perplexity and Google AIO
LinkedIn appeared in AI answers especially in Perplexity and Google AIO, and barely appeared in ChatGPT and Claude, in murmur’s weekly measurement. Google AIO is the AI-written summary that Google shows at the top of many searches, before the links. Perplexity is the AI that searches the web and cites its sources in the answer itself. Both cited LinkedIn articles and posts in dozens of the 90 questions, and ChatGPT in one.
murmur’s measurement asks the same 90 questions, round after round, and records the sources each AI shows. GEO, the work of making the company appear in the answers of ChatGPT and the other AI models, is the subject of the questions. The numbers apply to this market and to the four rounds of October and November 2026, and they are not a rule for each AI.
How much weight LinkedIn carries in AI answers, in murmur’s measurement
LinkedIn was a source in more than half of the questions in murmur’s measurement: 53 of the 90. The figure changes a lot from one AI to another, and LinkedIn’s robots.txt also treats each robot differently.
| AI | questions with LinkedIn among the sources (of 90) | what LinkedIn’s robots.txt says to its crawlers |
|---|---|---|
| Perplexity | 38 | closes the entire site to PerplexityBot and Perplexity-User |
| Google AIO | 30 | closes the entire site to Google-Extended |
| ChatGPT | 1 | opens to OAI-SearchBot, with folders closed; closes the entire site to GPTBot and ChatGPT-User |
| Claude | none in the recorded sources | opens to Claude-SearchBot, with folders closed; closes the entire site to ClaudeBot and Claude-User |
Perplexity’s documentation writes, about Perplexity-User, the robot that opens the page when someone asks a question, that “Since a user requested the fetch, this fetcher generally ignores robots.txt rules.”
The opening in robots.txt and the measured citation moved in opposite directions in this measurement. ChatGPT and Claude, whose search crawlers can enter LinkedIn, barely cited it in the recorded sources. Perplexity, with both crawlers closed in the file, was the AI that cited LinkedIn the most. The measurement records the cited source, not the path Perplexity took to the page. Perplexity has its own guide, with the crawlers and the sources it cites: how to appear in Perplexity.
Article, post and the other pages: what from LinkedIn becomes a source
The LinkedIn article became a source more often than the post in murmur’s measurement. Of the 115 LinkedIn citations across the four rounds, 60 pointed to articles and 41 to posts. The other 14 citations pointed to other LinkedIn pages.
The article is the long text that LinkedIn publishes with its own title and its own URL. The post is the short publication in the feed. In the measurement, both entered the sources of Perplexity and Google AIO, and the article came out ahead.
The author’s profile also has a control that the owner sets. LinkedIn Help describes public profile visibility like this: “Control which sections of your profile can be displayed to people who aren’t connected to a LinkedIn account or can be viewed in search tools such as Google or Bing.” That control deals with search engines such as Google and Bing, and LinkedIn’s robots.txt separately decides what each AI bot may read.
In LinkedIn’s robots.txt, only the ChatGPT and Claude search bots get in, with folders closed
In LinkedIn’s robots.txt, read on 7 November 2026, only the ChatGPT and Claude search bots are allowed in, and even they have folders closed. The robots.txt is where each site writes, bot by bot, what may be read. AI bots, or crawlers, read pages for the AI companies, and OpenAI’s GPTBot is one of them.
LinkedIn’s robots.txt separates the bots by function. OAI-SearchBot, which fetches pages for ChatGPT’s answers, and Claude-SearchBot, which does the same for Anthropic’s Claude, are allowed in with the /public-profile/ folder closed. The training bots stay out of the entire site, with Disallow: /: OpenAI’s GPTBot, Anthropic’s ClaudeBot and Google-Extended. Also kept out are ChatGPT-User and Claude-User, which open a page when someone asks in the conversation, and Perplexity’s two bots, PerplexityBot and Perplexity-User.
LinkedIn’s setup shows that a site can let search in and block training at the same time. Each bot has its own line in the file, and the same AI company can have one bot allowed and another blocked. How robots.txt works on any site, and what it changes for AI models, is in the guide about llms.txt, robots.txt and what works to be cited by AI.
LinkedIn in the studies that measured outside Brazil, each with its own scope
Outside Brazil, the 2026 studies also saw LinkedIn strong in Perplexity, each with a different scope. Bill Widmer, in a study published on 2 September 2026 by Orbit Media, a company that sells website creation with an SEO focus, counted 13,184 citations in 1,765 answers to 72 purchase questions about three brands that sell to businesses. In Perplexity, LinkedIn was the most cited site in that study, with 374 citations, ahead of YouTube and Reddit. The author declares Semrush as a client, and the data goes up to 23 August 2026.
Ahrefs, a company that sells SEO tools, measures on another scale. In the list of the 50 sites most cited by Perplexity, updated on 3 November 2026, linkedin.com came 20th, with 1.1% of the combined citations of those 50 sites. The calculation is based on questions on all subjects asked in the United States.
The two numbers are not added together or compared. The Orbit Media one comes from purchase questions about three brands, and the Ahrefs one from all subjects. Both are about Perplexity, and neither makes LinkedIn the source most cited by all AI models. The weight of LinkedIn, YouTube and the other networks in ChatGPT’s sources in Brazil, month by month, is in the guide about the sites that influence ChatGPT in Brazil.
How to publish on LinkedIn so the AI uses the text, in five steps
For the AI to use the LinkedIn text, publish articles that answer the buyer’s question right at the start and make the profile and the company page easy to recognize. That is the kind of text that became a source in murmur’s measurement, and a page on the website takes the same subject to ChatGPT and Claude, which barely cited LinkedIn.
- Write the article with the answer at the beginning. The first sentence of the article answers the question, with the company name and the main fact written in the text.
- Put the buyer’s question in the title. The article title uses the words the buyer types, not the technical name of the service.
- Keep the profile visible in search engines. In public profile visibility, choose which sections can appear in search tools such as Google and Bing.
- Give the company page the same name as the website. The LinkedIn company page uses the same name and the same description as the website, so the AI does not confuse the brand with another.
- Publish the same subject on a page of the website. The website page carries the same answer to ChatGPT and Claude, which barely cited LinkedIn in the measurement.
The answer at the beginning of the text applies to the LinkedIn article as it does to the website page, and how to write each one is in the guide on how to write a text with the answer at the beginning and numbers with sources. LinkedIn is an off-site channel, like Reddit and Quora, which have their own guide: Reddit and Quora to appear in AI answers. Video, which is another off-site door, is in the guide on why YouTube appears in AI answers.
The same question about another network, Instagram, has a similar answer in Google AIO: the Google summary used mostly the written caption of the post, in murmur’s study of social networks, as the guide on what Google’s AI Overview uses from an Instagram post shows.
In ChatGPT, the door is the website page: the Fly Vet case study
In murmur’s measurement, with questions about GEO, ChatGPT cited LinkedIn in 1 of 90. For these questions, the company’s answer also needs to be on a page of the website, as Fly Vet did.
Fly Vet is a veterinary marketing agency with its own technology and a murmur client. Fly Vet’s first content for AI models went live on 22 May 2026, and the measurement guided the production of the website. On 2 September 2026, ChatGPT recommended Fly Vet in first place in 30% of the questions about hiring a veterinary marketing agency. None of these questions mentions Fly Vet by name.
murmur places the brand inside AI answers in continuous work, with a person dedicated to the brand from the diagnostic onward. The specialist dives into the company and the sector, sees who the AI models recommend in place of the brand and closely follows the effect of each piece of content on the answers. The same questions return every week to ChatGPT, Claude, Perplexity and Google AIO, and the numbers about LinkedIn in AI answers came from this measurement by murmur itself.
Frequently asked questions
Article or LinkedIn post: which appears more in AI?
The article appeared more than the post in murmur’s measurement: of the 115 LinkedIn citations in the four rounds, 60 were articles and 41 were posts. Both appeared mostly in the sources of Perplexity and Google AIO. The article has its own title and URL, and therefore can answer a whole question on a single page.
Does ChatGPT read LinkedIn?
LinkedIn’s robots.txt lets OAI-SearchBot, ChatGPT’s search bot, into the site, with the /public-profile/ folder closed. The same file closes the entire site to GPTBot, OpenAI’s training bot, and to ChatGPT-User, which opens a page when someone asks in the conversation. In murmur’s measurement, with questions about GEO, ChatGPT cited LinkedIn in 1 of 90. For these questions, the company’s answer also needs to be on a page of the website.
Does Claude cite LinkedIn articles?
In Claude’s recorded sources in murmur’s measurement, LinkedIn did not appear in any of the 90 questions. LinkedIn’s robots.txt lets Claude-SearchBot, Claude’s search bot, in, with the /public-profile/ folder closed, and closes the entire site to ClaudeBot, the training bot, and to Claude-User. The company’s answer for Claude, therefore, is better placed on a page of its own website.
How do you make a LinkedIn profile visible in search engines?
A LinkedIn profile is visible in search engines through public profile visibility, the control that LinkedIn Help describes. In it, the owner chooses which sections of the profile can be shown to people who are not connected to LinkedIn or viewed in search tools such as Google and Bing. This control deals with search engines, and LinkedIn’s robots.txt separately decides what each AI bot can read.
Does Perplexity cite LinkedIn a lot outside Brazil?
It depends on who measured and what was asked. In Bill Widmer’s study, published by Orbit Media on 2 September 2026, with 72 buying questions about three brands that sell to companies, LinkedIn was the site most cited by Perplexity, with 374 citations. In the Ahrefs list updated on 3 November 2026, with questions on all subjects in the United States, linkedin.com came in 20th place among the 50 sites most cited by Perplexity. In murmur’s measurement, Perplexity cited LinkedIn in 38 of the 90 questions.
Where to start
Before the next LinkedIn article, it is worth seeing how AI models answer your company’s market questions today. Just enter the site in murmur’s free diagnostic: with no sign-up, in less than two minutes, the AI models’ answers to your market’s questions appear, and the full report follows by email. murmur’s sector studies show which brands the AI models recommended in the sectors murmur measured.
Read also
- Perplexity: its bots and the sources it cites
- Sites that influence ChatGPT in Brazil: company sites lead, and Reddit fell
- Reddit and Quora to appear in AI models: where they still work and how to avoid getting banned
- llms.txt: What It Is, What It Is For, and Whether It Works to Get Cited by AI
References
- murmur, weekly measurement with 90 questions about GEO and GEO agencies in ChatGPT, Claude, Perplexity and Google AIO, in four rounds from 12 October to 2 November 2026; no published page.
- LinkedIn, the site’s robots.txt. OAI-SearchBot and Claude-SearchBot with “Disallow: /public-profile/”; GPTBot, ClaudeBot, Google-Extended, ChatGPT-User, Claude-User, PerplexityBot and Perplexity-User with “Disallow: /”. https://www.linkedin.com/robots.txt (accessed on 7 November 2026).
- LinkedIn, LinkedIn Help, public profile visibility, undated page. The owner controls which sections of the profile appear to people who are not connected and in search tools such as Google or Bing. https://www.linkedin.com/help/linkedin/answer/a518980 (accessed on 7 November 2026).
- Perplexity, “Perplexity Crawlers”, documentation, undated page. “Since a user requested the fetch, this fetcher generally ignores robots.txt rules.” (about Perplexity-User). https://docs.perplexity.ai/docs/resources/perplexity-crawlers (accessed on 6 October 2026).
- Bill Widmer, Orbit Media, “I Tracked 13,184 AI Citations Across 4 LLMs: Here’s What I Learned”, 2 September 2026. On Perplexity, “Its all-time favorites in my data are LinkedIn (374 citations), YouTube (248), and Reddit (232).”; 72 buying questions about three brands that sell to companies, data up to 23 August 2026; the author declares Semrush as a client. https://www.orbitmedia.com/blog/ai-citation-sources/ (accessed on 7 October 2026). Orbit Media sells website creation with a focus on SEO.
- Ahrefs, list of the 50 sites most cited in Perplexity, updated on 3 November 2026, United States, all subjects. linkedin.com in 20th place, with 1.1% of the combined citations of the 50 sites. https://ahrefs.com/blog/most-cited-domains-perplexity/ (accessed on 7 November 2026). Ahrefs sells SEO tools.