SolCrys Logo

Citation & Source Influence

The social web each AI engine sees: crawler access and citations by platform (October 2026)

Each AI answer engine sees a different slice of the social web, set by which crawlers each platform admits and which licensing deals it signs. Most large social platforms admit Googlebot but block the search crawlers of OpenAI and Anthropic, and most block Perplexity's too, so social content reaches AI answers mostly through Google and through the platforms' own assistants, such as Grok on X. In the answers SolCrys tracks, Google AI Overviews cites YouTube most often among social platforms, then LinkedIn, Reddit, Facebook, and Quora, while ChatGPT rarely cites any. This page maps the crawler rules of 13 platforms, the deals behind them, and where to put social effort by engine, as of Oct. 16, 2026.

Updated

Questions this guide answers

  • Does ChatGPT cite TikTok or Instagram posts?
  • Which social media platforms do AI answer engines cite?
  • Why does Google AI Overviews cite social media more than ChatGPT?
  • Can ChatGPT or Perplexity read posts on X?
  • Which social platform matters most for AI search visibility?

Direct answer

Social content reaches AI answers mostly through Google and through the platforms' own assistants. As of Oct. 16, 2026, X, TikTok, Threads, Pinterest, and Quora let Googlebot crawl public posts (all but Quora admit Bingbot too) but block every search crawler and user-triggered fetcher from OpenAI, Anthropic, and Perplexity. Instagram and Facebook block OpenAI's and Anthropic's search crawlers and all user-triggered fetchers, yet admit their training crawlers and PerplexityBot. Reddit blocks every crawler in its public robots.txt and licenses its content to Google and OpenAI instead.

The answers follow the access. In a 30-day sample of AI answers SolCrys collected across B2B and consumer categories, ending Oct. 16, 2026, Google AI Overviews leaned on social platforms far more than any other engine we track, citing YouTube most often, then LinkedIn, Reddit, Facebook, and Quora. Gemini cited YouTube and Reddit but almost never the platforms that block Google-Extended, such as LinkedIn. ChatGPT rarely cited any social platform, Perplexity's social citations were mostly LinkedIn, and TikTok almost never appeared.

X belongs to xAI, whose Grok searches X posts natively, and Facebook's own AI Mode answers from public posts. So allocate by engine: social is a strong lever for Google AI Overviews, a narrower one for Gemini and Perplexity, and a weak one for ChatGPT, where your own site and third-party pages carry the answer.

Who can crawl what: 13 platforms, checked Oct. 16, 2026

We fetched each platform's robots.txt on Oct. 16, 2026, and read it per RFC 9309: a crawler without its own group follows the catch-all (*) group, and the most specific matching rule wins. "Allowed" means public post pages are open. OpenAI's column covers OAI-SearchBot and ChatGPT-User; Perplexity's, PerplexityBot and Perplexity-User; Anthropic's, Claude-SearchBot and Claude-User. Training crawlers are noted where they differ.

PlatformGoogle (Googlebot)OpenAI search and user botsPerplexity botsAnthropic botsAlso worth knowing
XAllowedBlockedBlockedBlockedAmong search and AI crawlers, only Googlebot and Bingbot get in; Google-Extended is blocked by name. Owned by xAI, whose Grok searches X directly.
TikTokAllowedBlocked by nameBlocked by nameBlocked by nameNames and blocks 25 agents, including Google-Extended, Gemini-Deep-Research, and Google-NotebookLM.
InstagramAllowedBlocked (GPTBot, the training crawler, is allowed)PerplexityBot allowed; Perplexity-User blockedBlocked (ClaudeBot, the training crawler, is allowed)Owned by Meta; Google-Extended blocked. The file's header says automated collection needs Instagram's written permission.
FacebookAllowedBlocked (GPTBot allowed)PerplexityBot allowed; Perplexity-User blockedBlocked (ClaudeBot allowed)Same pattern and header as Instagram. Facebook search has its own AI Mode.
ThreadsAllowedBlockedBlockedBlockedOwned by Meta. Unlike Instagram and Facebook, it also blocks GPTBot, ClaudeBot, and Google-Extended.
YouTubeAllowedAllowed on watch pagesAllowed on watch pagesAllowed on watch pagesOwned by Google. No AI-specific rules, but caption endpoints are closed to every crawler.
LinkedInAllowedOAI-SearchBot allowed; ChatGPT-User and GPTBot blockedBlockedClaude-SearchBot allowed; Claude-User and ClaudeBot blockedOwned by Microsoft. Blocks every crawler it doesn't name; Google-Extended is blocked by name.
RedditBlocked in the public file; Google licenses Reddit content (2024)Blocked; OpenAI licenses Reddit content (2024)Blocked; Reddit is suing Perplexity over scrapingBlocked; Reddit sued Anthropic in 2025The public file disallows every crawler. Access runs on contracts.
QuoraAllowed on question pages; most single-answer URLs blockedBlocked by nameBlockedBlocked by nameGoogle-only: Bingbot isn't named, so the catch-all blocks it, while Google-Extended is allowed. The header bars AI training without a licensing contract.
MediumAllowedAllowed (GPTBot blocked)AllowedAllowed (ClaudeBot blocked)Blocks training crawlers only, and adds a License line pointing to its licensing terms.
SubstackAllowedAllowedAllowedAllowedNo AI-specific rules on substack.com or the publications we checked; each publication serves its own file.
PinterestAllowedBlockedBlockedBlockedAn allowlist of named crawlers that includes Bingbot but no OpenAI, Anthropic, or Perplexity agent; the catch-all blocks everything else, Google-Extended included.
GitHubAllowed on repository pages and filesAllowed on repository pages and filesAllowedAllowedOwned by Microsoft. Directory views (/tree/) and raw files are closed to most crawlers.

Reading the matrix: 5 patterns

The table shows policy as written, and 5 patterns matter for planning.

  • Google is the crawler nearly every platform admits. Every platform here except Reddit lets Googlebot reach public posts, and Reddit licenses its content to Google. That is why Google AI Overviews, which draws on Google's Search index, cites social posts most.
  • OpenAI and Anthropic are locked out of the big consumer platforms, and Perplexity nearly so. Among the consumer platforms that block OpenAI and Anthropic, only Instagram and Facebook admit PerplexityBot. LinkedIn admits OpenAI's and Anthropic's search crawlers but not Perplexity's, and YouTube, Medium, Substack, and GitHub are open to all 3.
  • Meta's files look inverted. Instagram and Facebook admit GPTBot and ClaudeBot, which collect training data, plus PerplexityBot, yet block OAI-SearchBot, Claude-SearchBot, and every user-triggered fetcher. Both files also say automated collection needs the platform's written permission.
  • Gemini tracks Google-Extended, not Googlebot. Google says Google-Extended governs grounding in Gemini Apps and "does not impact a site's inclusion in Google Search." In our sample, Gemini cites YouTube and Medium, which allow it, and Reddit, which licenses to Google, but almost never LinkedIn, Instagram, X, or TikTok, which block it. Quora is the exception: it allows Google-Extended, yet Gemini almost never cites it.
  • robots.txt states a policy, not proof of access. OpenAI says robots.txt rules "may not apply" to ChatGPT-User, and Perplexity says Perplexity-User "generally ignores robots.txt rules"; Anthropic says its bots honor them. Perplexity cites LinkedIn and, occasionally, X, though both block its crawlers.

What each engine cites from social platforms

In a 30-day sample of AI answers SolCrys collected across B2B and consumer categories, ending Oct. 16, 2026, the engines split along the access lines above. We collect Perplexity with a recency filter that favors new pages, so read its row as a lighter signal.

EngineSocial platforms it citesRarely or neverWhat it means
Google AI OverviewsYouTube most often, then LinkedIn, Reddit, Facebook, and Quora; Instagram and Medium less oftenX and TikTok almost never, though both admit GooglebotThe most social-heavy engine we track. Social is a real lever here.
GeminiYouTube and Reddit, and Medium to a lesser degreeFacebook rarely; LinkedIn, Instagram, and Quora almost never; X and TikTok not seenPlatforms that block Google-Extended are nearly absent.
ChatGPTRarely any; the few are mostly LinkedIn, then Reddit and GitHubYouTube almost never; Facebook, Instagram, X, TikTok, and Quora not seenYour own site and third-party pages carry ChatGPT, not social posts.
PerplexityMostly LinkedInX occasionally (almost never on other engines); TikTok not seenLinkedIn is the platform to watch, despite its block on Perplexity's crawlers.
ClaudeNone seen so farEvery platform, in a smaller sample than the other enginesToo early to plan around.

What independent research shows

Neutral studies agree that Google's AI answers lean on a few social platforms and that ChatGPT leans on them less. In March 2025 browsing data from 900 U.S. adults, Pew Research Center found Wikipedia, YouTube, and Reddit were the most-cited sources in Google's AI summaries, together 15% of sources.

A Washington University in St. Louis study in the proceedings of the 2026 ACM Internet Measurement Conference ran 55,393 trending U.S. queries from March 13 to April 21, 2026. The most-cited hostname in AI Overviews was youtube.com (5.49%), with facebook.com (3.68%) and instagram.com (3.65%) right behind Wikipedia, and Reddit under 1%. Social and video platforms supplied 14.2% of AI Overview references, against 49.9% of first-page results, from 9.3% in Climate to 28.9% in Beauty & Fashion. In our sample, which includes B2B categories, LinkedIn ranked second.

For ChatGPT, researchers at Ruhr University Bochum and the Max Planck Institute for Software Systems (Kirsten et al.) found OpenAI's GPT models relied heavily on company sites and encyclopedias, "with less reliance on Social Media and User Forums," while Google's organic results drew up to 35% from such sites.

The 2026 licensing and ownership layer

Access to social content is increasingly licensed, litigated, or owned, and each deal changes what a single engine sees.

  • Reddit licenses and litigates. It licensed its content to Google in February 2024 and to OpenAI in May 2024, sued Anthropic in June 2025, and on July 31, 2026, won a ruling that lets the main claims of its scraping suit against Perplexity and SerpApi proceed. Licensed access isn't the same as citation: ChatGPT all but stopped citing Reddit in August 2026 (our citation study).
  • Reddit's Google deal is in renewal talks. CNBC reported on July 22, 2026, citing The Wall Street Journal, that the $60 million-a-year deal is ending soon and Reddit has discussed shutting off Google's AI access; we found no announced outcome by Oct. 16. Reddit also ends RSS feeds on Nov. 13, 2026, and public API access by March 2027.
  • X belongs to an AI company. xAI acquired X in March 2025, and SpaceX acquired xAI in February 2026. Since June 2025, X's developer terms bar using X content "to fine-tune or train a foundation or frontier model," while Grok's X Search tool searches X posts and answers with citations.
  • Meta, Yelp, and Wikipedia. On June 15, 2026, Meta unveiled an AI Mode for Facebook search that answers from public posts, including Groups and Reels. Yelp licenses reviews, photos, and business information to OpenAI so ChatGPT can link to them (reported July 23, 2026). On Jan. 15, 2026, Wikimedia named Amazon, Meta, Microsoft, Mistral AI, and Perplexity as Wikimedia Enterprise partners, alongside Google.
  • Cloudflare changed the rules for sites behind it. Since Sept. 15, 2026, its Block option for AI training also stops Googlebot, Bingbot, and Applebot, and Pay Per Use entered beta on Sept. 30 (details).

Where to put social effort, by engine

Start from your buyers: who is asking, what they need done, and at which buying stage. Those stay stable while engines shift, so map each persona and intent to its engine, then to the platforms that engine reads. In our LinkedIn and YouTube analyses, category questions ("best X for Y") drew mostly on independent voices, and brand questions on a brand's own posts and on others' posts about it.

  • If your buyers use Google AI Overviews, social is a real lever: YouTube for how-to, comparison, and review questions, LinkedIn for B2B, Facebook and Instagram for consumer questions, and Reddit and Quora where first-hand experience counts. Keep posts public and indexable. Google calls optimizing for its generative AI features "still SEO" (Google's guide), so treat this as an operating layer on top of SEO foundations.
  • If they use Gemini, YouTube, Reddit, and Medium reach it; LinkedIn, Instagram, X, and TikTok posts mostly don't.
  • If they use ChatGPT, don't count on social posts. Put the answer on your own site, keep OAI-SearchBot and ChatGPT-User allowed, and earn coverage on third-party pages ChatGPT can crawl, such as press and review sites (how to get cited by ChatGPT). LinkedIn, which admits OpenAI's search crawler, is the social platform ChatGPT cites most in our sample.
  • If they use Perplexity, publish substantive, public LinkedIn posts, and keep your own site open to PerplexityBot and Perplexity-User.
  • If they use Grok, post natively on X: Grok searches X posts directly and cites them. The same post does little for ChatGPT or Claude, whose crawlers X blocks.
  • If they live on Facebook, keep your public posts and Group answers accurate, because Facebook's AI Mode answers from them.

How to measure it

Measure from the answers, engine by engine; no social platform's analytics counts AI citations of your posts. Search Console's platform properties for Instagram, TikTok, X, and YouTube cover Search, Discover, and News, and Google's guide to them doesn't mention AI Overviews or AI Mode (what the free reports show).

Fix a prompt set by persona, intent, and buying stage, built from search demand, the questions people ask around it, and community discussion where available (our Golden Prompt Set). Run it on each engine repeatedly and log every cited social URL with its platform, its author (your company, your people, or independent), and whether it names you. Judge a change across several runs, not one (how many runs you need), and before crediting a post, check whether untouched prompts on the same engine moved too.

What may change

Everything here is as of Oct. 16, 2026. Watch 4 dated items: the outcome of Reddit's Google renewal talks, reported July 22, 2026; Reddit's RSS shutdown on Nov. 13, 2026, and API changes through March 2027; Google's work to name creators in AI responses (May 6 and Oct. 15, 2026); and edits to any platform's robots.txt, which can land any day.

How SolCrys helps

SolCrys runs your buyer prompts on a schedule on ChatGPT and, on paid plans, Gemini, Google AI Overviews, Perplexity, and Claude, and records every cited URL, social posts included, so you can see which platforms each answer leaned on and whether those posts mention you or competitors. Workspace Deep Analysis can recommend a platform post where one would close a gap, with governance guidance attached: post from your own account, disclose your affiliation, use first-hand data, and no engagement pods.

That is Measure → Diagnose → Execute → Verify applied to social: your team writes and approves the post, and the frozen prompt set reruns on the same engines so you can judge across runs whether answers moved. SolCrys doesn't track AI Mode, Copilot, Grok, or Meta AI. Start Free with 10 ChatGPT prompts and 1 content audit, free, with email verification and no credit card.

Sources

FAQ

Does ChatGPT cite TikTok or Instagram?

Not in the answers SolCrys tracks. TikTok blocks OpenAI's crawlers by name, and Instagram's robots.txt admits only GPTBot, OpenAI's training crawler, while blocking OAI-SearchBot and ChatGPT-User. ChatGPT rarely cites any social platform; the few social citations we see are mostly LinkedIn, which admits OpenAI's search crawler.

Why does Google AI Overviews cite social posts more than ChatGPT does?

Access, mostly. Nearly every social platform admits Googlebot, while most block OpenAI's search crawler; Google also owns YouTube and licenses Reddit's content. AI Overviews draws on Google's Search index, which includes those posts. Even so, AI Overviews cites social far less than the blue links do: in a 2026 Washington University in St. Louis study, social and video platforms were 14.2% of AI Overview references versus 49.9% of first-page results.

Can Perplexity read X posts?

X's robots.txt blocks PerplexityBot and Perplexity-User, along with every search and AI crawler except Googlebot and Bingbot. Perplexity says Perplexity-User generally ignores robots.txt, and X posts do appear occasionally in Perplexity's answers in our sample. The engine that reads X natively is Grok, through its X Search tool.

Which social platform matters most for AI search?

It depends on the engine your buyers use. YouTube matters most for Google AI Overviews and Gemini, LinkedIn for Perplexity and for B2B questions in Google AI Overviews, and X for Grok. For ChatGPT, no social platform matters much; your own site and third-party pages do.

Why doesn't Gemini cite LinkedIn or Instagram?

Both block Google-Extended, which Google documents as the control for grounding in Gemini Apps, and in our sample Gemini almost never cites either. Google AI Overviews relies on Googlebot instead, which both platforms allow.

Free ChatGPT visibility check

See where AI answers skip your brand — then fix it, free

Start a free workspace with your domain: 10 buyer-intent prompts through ChatGPT show where you are mentioned, cited, or skipped, and who gets recommended instead. A free content audit in the same workspace hands you the first fix to ship.

Start Free

Free · No credit card · About 5 minutes

Related guides

Community & UGC

Reddit AEO Playbook

A 6-step playbook for Reddit as an AEO surface - weekly cadence, thread-authoring patterns, disclosure norms, a 12-week build plan - re-underwritten after ChatGPT stopped showing Reddit citations in August 2026. What Reddit still buys you, and on which engines.

AI Engine Optimization

How to Optimize for Google AI Overviews & AI Mode

How Google's AI Overviews and conversational AI Mode are built from your index, the inheritance signals that decide which pages get cited, and the optimization plays that actually move position.