Your business doesn't appear in ChatGPT because ChatGPT's answer layer only cites pages its crawlers can fetch, parse, and confidently attribute — and most sites fail on at least one of those three.
ChatGPT's live web results are gathered by two OpenAI crawlers — GPTBot and OAI-SearchBot — and, for ChatGPT Search, augmented by Bing's index. If your robots.txt blocks any of those user agents, you are invisible on that channel by design.
Even when crawling is allowed, the answer layer needs machine-parseable content. Pages rendered entirely on the client (JavaScript-only) often ship an empty HTML shell to crawlers; Google's own guidance on JavaScript SEO spells out why server-rendered HTML is safer for retrieval. Most AI crawlers are lighter-weight than Googlebot and are less forgiving.
Finally, the model has to decide who to credit. That job is made easy by explicit entity signals — QAPage or Article JSON-LD with an author Person node, sameAs links to authoritative profiles, and a stable canonical URL. Without those, the answer engine may still use your content while attributing it to someone else who marked it up.
Fix the three in order: allow the AI crawlers, ship real HTML on first byte, mark up the answer + author. Then request a fresh crawl (Google Search Console for AI Overviews, Bing Webmaster + IndexNow for ChatGPT Search).