Why ChatGPT Ignores My Site — Root Causes | AiVIS.biz
ChatGPT with browsing and SearchGPT use GPTBot to crawl and extract content. If GPTBot cannot access, parse, or attribute your page, ChatGPT will cite someone else — or hallucinate.
How ChatGPT decides what to cite
When ChatGPT generates a response that draws on web content, it relies on GPTBot's crawl corpus and real-time browsing results. The model extracts fragments, compresses them, and reconstructs an answer. Pages that provide clear, atomic claims with structured metadata get cited. Pages that are vague, unstructured, or inaccessible get skipped.
ChatGPT does not 'ignore' your site out of preference. It ignores your site because the extraction pipeline failed at one or more stages.
Check if GPTBot can access your site
Open your robots.txt file and search for GPTBot. If you see 'Disallow: /' under the GPTBot user-agent, ChatGPT cannot crawl any page on your site. Many CDN security configurations (Cloudflare Bot Fight Mode, Sucuri WAF) also block GPTBot by default.
An AiVIS.biz audit checks GPTBot access as part of the technical SEO dimension and flags blocked crawlers as high-confidence findings.
Ensure your content is extractable
Even if GPTBot can access your page, it still needs to extract usable content. Client-side-only rendering produces an empty shell. Missing JSON-LD means the model cannot identify the publisher. Vague headings make it harder for the model to segment and attribute claims.
AiVIS.biz audits all six extraction dimensions and provides specific, code-level fix recommendations for each failure.
Frequently Asked Questions
- Does blocking GPTBot affect ChatGPT answers about my brand?
- Yes. If GPTBot is blocked, ChatGPT cannot crawl your site for real-time information. It may still have older training data, but real-time answers and citations will exclude your content.
- Can I allow GPTBot without allowing all AI crawlers?
- Yes. You can selectively allow GPTBot in robots.txt while blocking other crawlers. Each AI crawler has a different user-agent string.