{"id":353,"date":"2026-11-14T07:00:00","date_gmt":"2026-11-14T07:00:00","guid":{"rendered":"https:\/\/livementions.com\/blog\/?p=353"},"modified":"2026-10-19T11:52:45","modified_gmt":"2026-10-19T11:52:45","slug":"where-does-chatgpt-get-its-sources","status":"publish","type":"post","link":"https:\/\/livementions.com\/blog\/where-does-chatgpt-get-its-sources\/","title":{"rendered":"Where Does ChatGPT Get Its Sources? What Your Category&#8217;s Cited Domains Look Like"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">Ask ChatGPT for the best option in your category and it may answer with a short list and a row of citations. Where does ChatGPT get its sources for that answer? OpenAI says it searches the web when a question benefits from current information, and the response may include citations you can open. Your category has its own pattern of cited domains, and a Sources page can show it.<\/p>\n\n\n<h2 class=\"wp-block-heading\" id=\"h-where-does-chatgpt-get-its-sources-according-to-openai\">Where does ChatGPT get its sources, according to OpenAI?<\/h2>\n\n\n<p class=\"wp-block-paragraph\">OpenAI publishes four facts that bear on this question, and each comes from its own pages.<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><strong>ChatGPT searches the web when it helps.<\/strong> The help article on <a href=\"https:\/\/help.openai.com\/en\/articles\/9237897-chatgpt-search\" target=\"_blank\" rel=\"nofollow noopener noreferrer\">searching the web with ChatGPT<\/a> says ChatGPT may search automatically when a question would benefit from current information, and that you can start a search manually.<\/li>\n\n<li><strong>Queries go to search providers.<\/strong> The same article says ChatGPT search sometimes partners with other search providers and typically rewrites a question into one or more targeted queries that it sends to them. The <a href=\"https:\/\/openai.com\/index\/introducing-chatgpt-search\/\" target=\"_blank\" rel=\"nofollow noopener noreferrer\">announcement of ChatGPT search<\/a> adds that it also uses content provided directly by partners.<\/li>\n\n<li><strong>Sites can opt in through robots.txt.<\/strong> The <a href=\"https:\/\/platform.openai.com\/docs\/bots\" target=\"_blank\" rel=\"nofollow noopener noreferrer\">OpenAI crawler overview<\/a> says OAI-SearchBot is used to surface websites in search results in ChatGPT&#39;s search features. Sites that opted out of it will not be shown in ChatGPT search answers, though they can still appear as navigational links.<\/li>\n\n<li><strong>Responses may include citations.<\/strong> You can select a citation to open its source, or select Sources, when available, to see cited sources and other relevant links. OpenAI adds that citations can be incomplete, outdated or incorrect.<\/li>\n<\/ul>\n\n\n<p class=\"wp-block-paragraph\">None of these pages lists how sources are ranked against each other. OpenAI describes where pages can come from and how they are shown, and it does not publish a formula for which page wins. Any claim that ChatGPT prefers one kind of site is therefore a measurement to make, not a rule to assume. Perplexity works in the same visible way: its help center says <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/10352895-how-does-perplexity-work\" target=\"_blank\" rel=\"nofollow noopener noreferrer\">each answer includes numbered citations that link to the original sources<\/a>.<\/p>\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1200\" height=\"675\" src=\"https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-steps.png\" alt=\"Where does ChatGPT get its sources: four steps from a question to cited pages: ChatGPT may search the web, sends queries to search providers, surfaces pages from sites that allow OAI-SearchBot, and cites sources in the answer\" class=\"wp-image-340\" title=\"\" srcset=\"https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-steps.png 1200w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-steps-300x169.png 300w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-steps-1024x576.png 1024w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-steps-768x432.png 768w\" sizes=\"auto, (max-width: 1200px) 100vw, 1200px\" \/><\/figure>\n\n\n<h2 class=\"wp-block-heading\" id=\"h-why-can-39-t-one-answer-show-where-chatgpt-can-get-its-sources\">Why can&#39;t one answer show where ChatGPT can get its sources?<\/h2>\n\n\n<p class=\"wp-block-paragraph\">One answer is one sample. Each AI provider can return different citations for the same question: one may cite your website, another may cite a third party, and another may return no citation at all. The Help article on <a href=\"https:\/\/livementions.com\/help\/sources\/why-source-data-differs-between-models\/\">why source data differs between models<\/a> says LiveMentions shows the raw result each provider returned, so the differences come from the providers.<\/p>\n\n\n<p class=\"wp-block-paragraph\">Two more details matter when you read citations:<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li>A response may carry no source at all. The Help article on why a response may have no sources gives two cases: the provider returned no citations, or the response held no source LiveMentions could record.<\/li>\n\n<li>Citation dates are missing for some providers.<\/li>\n<\/ul>\n\n\n<p class=\"wp-block-paragraph\">A pattern across many tracked questions is more reliable than any single answer. That is the job of the Sources page. For the question-by-question view of a single citation, the guide to <a href=\"https:\/\/livementions.com\/blog\/ai-citation-tracking\/\">AI citation tracking<\/a> covers how to read it.<\/p>\n\n\n<h2 class=\"wp-block-heading\" id=\"h-what-does-the-sources-page-show-for-your-category\">What does the Sources page show for your category?<\/h2>\n\n\n<p class=\"wp-block-paragraph\">The <a href=\"https:\/\/livementions.com\/docs\/sources-page\/\">Sources page<\/a> answers a category question: which sites show up as sources across all your tracked prompts. The Docs say it shows the websites that appear most often, each with a percentage share of total citations. It adds a full table of every recorded citation: title, URL, AI platform, whether it was directly cited or just referenced, and the date. The Help center defines a source as a page or domain cited in an AI response.<\/p>\n\n\n<p class=\"wp-block-paragraph\">Since 30 September 2026 the page also breaks citations down by type. The changelog entry describes a split by domain type (corporate, news, community, social, academic and more) and by citation type (blog article, documentation, listicle, directory and more). The numbers below come from one tracked project, which is a real project in LiveMentions and not a market study.<\/p>\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1348\" height=\"470\" src=\"https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-types.png\" alt=\"Domain Types and Citation Types tables from one tracked project: corporate sites hold 90 percent of 224 citations, and website pages, blog articles and listicles lead the citation types\" class=\"wp-image-341\" title=\"\" srcset=\"https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-types.png 1348w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-types-300x105.png 300w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-types-1024x357.png 1024w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-types-768x268.png 768w\" sizes=\"auto, (max-width: 1348px) 100vw, 1348px\" \/><\/figure>\n\n\n<h3 class=\"wp-block-heading\" id=\"h-what-do-domain-types-and-citation-types-look-like-in-a-real-project\">What do domain types and citation types look like in a real project?<\/h3>\n\n\n<p class=\"wp-block-paragraph\">In that project 224 citations were found. The two tables split the same 224 citations in two ways, by the kind of site and by the kind of page.<\/p>\n\n\n<figure class=\"wp-block-table\"><table class=\"has-fixed-layout\"><thead><tr>\n<th>Citation type<\/th>\n<th>Share<\/th>\n<th>Citations<\/th>\n<\/tr>\n<\/thead><tbody><tr>\n<td>Website page<\/td>\n<td>42%<\/td>\n<td>94<\/td>\n<\/tr>\n<tr>\n<td>Blog article<\/td>\n<td>30%<\/td>\n<td>67<\/td>\n<\/tr>\n<tr>\n<td>Listicle<\/td>\n<td>18%<\/td>\n<td>41<\/td>\n<\/tr>\n<tr>\n<td>Directory<\/td>\n<td>4%<\/td>\n<td>9<\/td>\n<\/tr>\n<tr>\n<td>Documentation<\/td>\n<td>4%<\/td>\n<td>8<\/td>\n<\/tr>\n<tr>\n<td>Social media post<\/td>\n<td>1%<\/td>\n<td>3<\/td>\n<\/tr>\n<tr>\n<td>Research<\/td>\n<td>0%<\/td>\n<td>1<\/td>\n<\/tr>\n<tr>\n<td>News<\/td>\n<td>0%<\/td>\n<td>1<\/td>\n<\/tr>\n<\/tbody><\/table><\/figure>\n\n\n<p class=\"wp-block-paragraph\">By domain type, corporate sites hold 90% (202 citations), commercial sites 4% (9), editorial sites 4% (9), social sites 1% (3) and academic sites 0% (1). Two readings follow, and both are facts about this project&#39;s answers, not about every category.<\/p>\n\n\n<ol class=\"wp-block-list\">\n<li><strong>Articles carry nearly half.<\/strong> Blog articles and listicles together are 108 of the 224 citations, about 48%.<\/li>\n\n<li><strong>Community and academic sources are rare here.<\/strong> Social and academic sites together hold 4 citations, and no community type appears in the domain table.<\/li>\n<\/ol>\n\n\n<p class=\"wp-block-paragraph\">Your category will not look the same. A category with active forums or a research base would show a different split, and that difference is what the page exists to reveal.<\/p>\n\n\n<h2 class=\"wp-block-heading\" id=\"h-how-do-you-see-whether-cited-pages-name-your-brand\">How do you see whether cited pages name your brand?<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Cited pages matter most when they name you. In the changelog entry of 30 September 2026, two stats were added to the Sources page: Mention Share and Brand Mentions. They show how often AI answers cite pages that actually mention your brand. Mention Share is the percentage of AI-cited documents that mention the brand. Brand Mentions is the total count of citations that mention the brand, out of all citations found.<\/p>\n\n\n<p class=\"wp-block-paragraph\">In the project above the figures read 11% and 24 out of 224. The arithmetic checks out: 24 of 224 is about 11%. It also means about nine in ten cited documents do not name the brand. The changelog adds that LiveMentions checks the real page before it <a href=\"https:\/\/livementions.com\/blog\/ai-brand-mention-detection\/\">marks a citation as a mention<\/a>, so a cited page that never names the brand is not counted.<\/p>\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1344\" height=\"138\" src=\"https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-mentions-fixed.png\" alt=\"Mention Share and Brand Mentions cards on the Sources page, showing a Mention Share of 11% and 24 brand mentions out of 224 citations found\" class=\"wp-image-429\" title=\"\" srcset=\"https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-mentions-fixed.png 1344w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-mentions-fixed-300x31.png 300w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-mentions-fixed-1024x105.png 1024w, https:\/\/livementions.com\/blog\/wp-content\/uploads\/2026\/10\/P13-mentions-fixed-768x79.png 768w\" sizes=\"auto, (max-width: 1344px) 100vw, 1344px\" \/><\/figure>\n\n\n<p class=\"wp-block-paragraph\">Under these cards, the citations table marks each cited page as Mentioned or Not mentioned, and each source as Directly cited or Referenced. Read the Not mentioned rows as a to-do list: those are <a href=\"https:\/\/livementions.com\/blog\/citation-gap-analysis\/\">pages AI used that do not name you<\/a>.<\/p>\n\n\n<h2 class=\"wp-block-heading\" id=\"h-how-do-you-see-where-chatgpt-can-get-its-sources-for-your-own-category\">How do you see where ChatGPT can get its sources for your own category?<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Five steps show where ChatGPT can get its sources in your category. The order is this guide&#39;s method, built from the pages the Docs describe.<\/p>\n\n\n<ol class=\"wp-block-list\">\n<li><strong>Check the top cited domains.<\/strong> Note whether a few domains hold a large share or the list is flat. A flat list means many sites matter, and a concentrated one means a few do.<\/li>\n\n<li><strong>Read Domain Types.<\/strong> See which kinds of site the answers rely on, such as corporate, news, community, social or academic.<\/li>\n\n<li><strong>Read Citation Types.<\/strong> See which kinds of page the answers use. A heavy share of listicles shows that ranked best-of articles are a source type to study, and a heavy share of documentation shows the same for docs pages.<\/li>\n\n<li><strong>Check Mention Share.<\/strong> If the share is low, most cited pages are places AI used without naming you.<\/li>\n\n<li><strong>Take the Not mentioned pages to Gap analysis.<\/strong> It ranks those sources by gap score and shows which tracked competitors each page names instead, as the Gap analysis documentation explains.<\/li>\n<\/ol>\n\n\n<p class=\"wp-block-paragraph\">The citation table names the AI platform that used each source. ChatGPT, Claude, Gemini and Perplexity can each cite different sources for the same question, so a domain that leads one engine may not appear in another.<\/p>\n\n\n<h2 class=\"wp-block-heading\" id=\"h-which-sources-should-you-act-on-first\">Which sources should you act on first?<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Act first on a source type you can change. If your category&#39;s answers lean on blog articles and listicles, a comparison page or a listing request fits. If they lean on directories, complete your listings. If they lean on community threads, a useful answer in the thread fits. These are recommendations, not claims about how any engine ranks pages.<\/p>\n\n\n<p class=\"wp-block-paragraph\">Before you invest in content, <a href=\"https:\/\/livementions.com\/blog\/how-to-get-cited-by-chatgpt\/\">check that the engines can read your site at all<\/a>. OpenAI says sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, so a robots.txt rule that blocks it keeps a page out of those answers. The From gap to fix guide notes that many sites block AI crawlers by accident through old rules, plugins or security settings.<\/p>\n\n\n<h2 class=\"wp-block-heading\" id=\"h-frequently-asked-questions\">Frequently asked questions<\/h2>\n\n<div id=\"rank-math-faq\" class=\"rank-math-block\">\n<div class=\"rank-math-list \">\n<div id=\"faq-question-1\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">Where does ChatGPT get its sources?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>OpenAI says ChatGPT may search the web when a question would benefit from current information, may send rewritten queries to search providers it partners with, and may include citations in the response. Pages from sites that allow OAI-SearchBot can be surfaced in its search features.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-2\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">Does ChatGPT always show its sources?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>No. OpenAI says responses that use web search may include citations, and the Sources button appears when available. A response may also have no sources LiveMentions can record, as the Help center explains.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-3\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">Can I control whether ChatGPT can use my site as a source?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Partly. OpenAI says sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as navigational links. OpenAI recommends allowing OAI-SearchBot in robots.txt and allowing requests from its published IP ranges.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-4\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">Are the cited domains the same in every category?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>No. The Sources page shows the pattern for your own tracked questions, and the split by domain type and citation type differs from one project to another. The project in this guide is one example.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-5\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \">How is this different from tracking citations answer by answer?<\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Answer-by-answer tracking records what one answer cited. The Sources page adds up every tracked answer into shares by domain, domain type and citation type, so you can see the category pattern.<\/p>\n\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n\n<h2 class=\"wp-block-heading\" id=\"h-see-which-domains-your-category-39-s-answers-cite\">See which domains your category&#39;s answers cite<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Open the Sources page to see where an assistant can get its sources for your questions, then check Mention Share. LiveMentions is an AI brand visibility platform. It tracks how often, and how favourably, a brand appears in answers from ChatGPT, Claude, Gemini and Perplexity, compares it with competitors, shows which sources the AI cites, and suggests ways to improve (the Optimize feature). The <a href=\"https:\/\/livementions.com\/platform\/geo-insights\/\">GEO Insights page<\/a> shows the Sources view and Gap analysis together.<\/p>\n\n","protected":false},"excerpt":{"rendered":"<p>Ask ChatGPT for the best option in your category and it may answer with a short list and a row of citations. Where does ChatGPT get its sources\u2026<\/p>\n","protected":false},"author":1,"featured_media":346,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[31],"tags":[14,7,10],"class_list":["post-353","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-product","tag-ai-citation-tracking","tag-ai-search","tag-ai-visibility"],"_links":{"self":[{"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/posts\/353","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/comments?post=353"}],"version-history":[{"count":2,"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/posts\/353\/revisions"}],"predecessor-version":[{"id":519,"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/posts\/353\/revisions\/519"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/media\/346"}],"wp:attachment":[{"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/media?parent=353"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/categories?post=353"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/livementions.com\/blog\/wp-json\/wp\/v2\/tags?post=353"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}