How AI Tools Choose Sources: ChatGPT, Google and Perplexity

Many site owners ask why an AI answer cites one page and ignores another. This guide explains how ai tools choose sources, using only what OpenAI, Google and Perplexity say in their own documentation. Where the companies say nothing, I say so too.

How do AI tools choose sources in general?

AI search tools choose sources by running a web search for your question, reading the pages that come back, and citing the ones they use in the answer.

Some terms help here:

• Retrieval: the step where the tool fetches web pages to answer a question, instead of relying only on what the model learned in training.
• Citation: a link in the answer that points to the page a fact came from.
• Crawler: a bot that visits pages. Each tool has its own, such as OAI-SearchBot or PerplexityBot.
• robots.txt: a file on your site that tells crawlers what they may visit.
• Snippet: the short text preview a search engine can show from your page.
• Query fan-out: Google’s term for running several related searches to build one answer.

The key point is that access comes first. If a tool’s crawler cannot reach your page, the tool cannot cite it. After that, the companies describe goals such as relevance and authority, but not exact weights.

How does ChatGPT search pick and cite sources?

ChatGPT search picks sources by searching the web when a question needs fresh information, then showing inline citations to the pages it used.

OpenAI’s ChatGPT search help page says “ChatGPT will automatically search the web if your question might benefit from information on the web.” Users can also start a search by hand. The same page says answers include inline citations, and a Sources button lists the cited pages and other relevant links. It also names “third-party search providers such as Bing” as partners that may receive disassociated search queries.

For site owners, OpenAI’s crawler documentation is the most useful part. It says “OAI-SearchBot is used to surface websites in search results in ChatGPT’s search features.” It also says “sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers.” A separate bot, GPTBot, controls training use. OpenAI says each setting is independent, so you can block training and still appear in search.

OpenAI’s publisher FAQ says “any public website can appear in ChatGPT search.” It adds that ChatGPT includes utm_source=chatgpt.com in referral URLs, so you can track visits in analytics. OpenAI does not publish a ranking formula. My guide on how to get cited by ChatGPT covers practical steps.

How do Google AI Overviews and AI Mode choose sources?

Google AI Overviews and AI Mode choose supporting links from pages that are indexed and eligible to show a snippet in Google Search, often using several related searches at once.

Google’s AI features and your website page says both features may use “query fan-out,” which means issuing “multiple related searches across subtopics and data sources.” That is why an AI answer can cite pages that do not rank on page one for the exact query.

To be a supporting link, Google says a page must be “indexed and eligible to be shown in Google Search with a snippet.” It also says “there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.” In short, the same SEO basics apply. You can limit what appears using nosnippet, data-nosnippet, max-snippet or noindex.

Google also says AI Overviews appear only when its systems decide they are “additive to classic Search.” So not every query shows one. For more detail, read my guide on how to rank in Google AI Overviews.

How does Perplexity pick and cite sources?

Perplexity picks sources by searching the internet in real time and building an answer with numbered citations that link to the original pages.

Perplexity’s help centre says it gathers information “from authoritative sources like articles, websites, and journals.” It says “each answer includes numbered citations linking to the original sources.” Its crawler documentation says “PerplexityBot is designed to surface and link websites in search results on Perplexity” and is not used to train foundation models.

Perplexity also explains source labels. Some cited domains carry a Government, Academic or Trusted label. Its help page says “Perplexity rates the whole website a source comes from, not each individual page.” The review checks things like whether a site corrects mistakes, says who wrote each piece, and keeps news apart from ads and opinion. It also says “no label does not mean a site is low quality.” My Perplexity ranking guide goes further.

How do the three tools compare?

All three tools retrieve live web pages and cite them, but each uses its own crawler and access rules.

How can you make your pages easier for AI tools to cite?

You make pages easier to cite by first allowing each crawler, then writing clear, well-sourced answers that a tool can lift and link.

1. Check robots.txt. Make sure OAI-SearchBot, PerplexityBot and Googlebot are not blocked.
2. Confirm indexing. Use Search Console to check key pages are indexed and not set to nosnippet.
3. Answer early. Put a direct answer in the first sentence under each question heading.
4. Show who wrote it. Add a named author and sources, which matches what Perplexity says its review checks.
5. Track referrals. Filter analytics for utm_source=chatgpt.com and watch AI referral traffic.

Brand mentions on trusted sites can also help tools connect your name with a topic. Clear entities help too, as explained in my entity SEO guide.

If you want help with this work, my GEO services cover AI search visibility from crawler access to content.

Frequently asked questions