How to show up in AI answers in 2026: what ChatGPT, Perplexity, Google and Copilot actually document

The short answer

Appearing in AI answers takes three things the platforms document: their search crawler can reach your pages, the pages are in their index, and the pages say something worth citing. Google says AI Overviews and AI Mode need no special optimization: a page must be indexed and eligible for a snippet, on a site that hasn't opted out in Search Console. OpenAI and Perplexity document which crawler to allow. Microsoft gives the most concrete writing advice and a citation report with page-level detail and sampled grounding queries.

Nobody documents a shortcut, and nobody can promise you a citation. This guide uses only what Google, Microsoft, OpenAI, Anthropic and Perplexity publish, checked on 8 October 2026. Where we add our own method, we say so.

Google AI Overviews and AI Mode: it's still SEO

Google's AI features documentation says there are no additional requirements or special optimizations for AI Overviews and AI Mode. A page must be indexed and eligible to show in Search with a snippet. Clicks and impressions from these features count in the regular Performance report, under the Web search type.

On 15 May 2026, Google published a guide to optimizing for its generative AI features, last updated 10 July 2026. It explains that the features are built on Google's core ranking systems: they retrieve pages from the Search index and run related background searches, which Google calls query fan-out. From Google's side, optimizing for AI search is still SEO, and "AEO" or "GEO" advice should be judged like any other third-party SEO advice.

What Google asks for

• Non-commodity content. Google contrasts generic tip lists, built from common knowledge, with first-hand accounts and expert takes that go beyond it. Write about what you know from doing the work.
• No pages for every variation. Making separate pages for each phrasing or fan-out query, mainly to influence rankings or AI responses, violates the scaled content abuse policy.
• Structure for readers. Paragraphs, sections and headings that make a page easy to follow.
• Technical basics. Crawlable content, a good page experience, less duplication and JavaScript that doesn't hide content.
• Business details. Google Business Profile and Merchant Center data can appear in AI responses.

What Google says you can skip

The same guide lists things to ignore for Google Search: llms.txt and other AI-specific files, splitting content into small "chunks" (there is no ideal page length), rewriting for AI (its systems understand synonyms), special schema markup for AI (structured data still helps with rich results), and chasing inauthentic mentions. On 15 May 2026, Google also clarified that its spam policies apply to generative AI responses.

Check one Search Console setting

Search Console now has a Search generative AI control under Settings. Sites are included by default. Excluding a site removes its links and content from AI Overviews, AI Mode and Discover's generative AI features, usually within a few days. Google says the setting isn't a ranking or inclusion signal for the rest of Search. It reached all sites worldwide on 31 August 2026.

Two details matter. Google's guide lists being included in this setting as a condition of eligibility. And Google-Extended, the robots.txt token for Gemini, does not control AI Overviews; it covers Gemini training and grounding elsewhere. To limit what Google shows from a page, the documented controls are nosnippet, data-nosnippet, max-snippet and noindex.

Let the right crawlers in

Each AI platform separates the crawler that builds its search index from the one that collects training data, and from the agent that visits a page while answering a user. This table summarizes what each company documents:

Sources: Google, OpenAI, Anthropic, Perplexity. Anthropic says all three of its bots honor robots.txt, and each can be allowed or blocked separately.

For a business that wants to be found, allow every search crawler and user agent. Training crawlers are a separate decision with no documented effect on search inclusion. Then test with real requests: a bot-protection rule at your CDN can refuse requests that robots.txt allows. Our own CDN refused default Python and Perl clients until we changed a rule; the details are in what an agent-ready website needs.

ChatGPT and Perplexity: access is what they document

OpenAI recommends allowing OAI-SearchBot and its published IP ranges if you want to appear in ChatGPT search. Sites that opt out are left out of search answers, though they can still appear as navigational links. OpenAI also says ChatGPT-User is not used to decide whether content appears in search.

Perplexity recommends allowing PerplexityBot and requests from its published IP ranges, and says robots.txt changes can take up to 24 hours to apply.

That is where their documentation stops. Neither crawler page describes how answers choose sources, and we found no ranking guidance from either company that we could cite. So we don't repeat third-party theories here. The rest of this guide (useful pages, evidence, real mentions, measurement) applies to them as general practice, not as documented ranking factors.

Copilot and Bing: the clearest advice and a real citation report

On 10 February 2026, Microsoft opened AI Performance in Bing Webmaster Tools as a public preview. It reports citations of your site across Microsoft Copilot, Bing's AI-generated summaries and select partner integrations:

• Total Citations: how often your pages appear as sources in AI answers.
• Average Cited Pages: unique pages from your site cited per day.
• Grounding queries: a sample of the phrases the AI used to retrieve your content.
• Page-level citations: counts per URL, with trends over time.

On 16 June 2026, Bing added Intents, Topics, Citation Share and Compare. Bing describes Citation Share as an observational metric, not a ranking system, and says it doesn't represent traffic share. Neither post describes click data.

Bing's advice for being cited

Microsoft's launch post is the most specific guidance any platform gives:

• Go deeper in related topics to show expertise.
• Use clear headings, tables and FAQ sections.
• Support claims with examples, data and cited sources.
• Keep content current.
• Make text, images and video describe the same things consistently.
• Use IndexNow to report added, updated and removed pages.
• For local businesses, keep Bing Places details current.

IndexNow notifies Bing, Yandex, Seznam, Naver, Yep and Amazon at once; a submission to one is shared with all. Google is not on the list. You can submit new, updated and deleted URLs, including pages that now return 404 or 410. Submitting doesn't guarantee indexing.

One conflict worth knowing: Google deprecated FAQ rich results in May 2026 and removed their documentation in June (Search Central updates), while Bing still recommends FAQ sections. Write FAQs where readers actually have those questions, not for markup.

Why third-party mentions matter

Google's guide says its AI features can show what's being said about products and services across the web, including blogs, videos and forum discussions. It also says chasing inauthentic mentions isn't as helpful as it might seem: its AI features depend on both the core ranking systems, which focus on quality, and the separate systems that block spam.

We see the practical side on our own name. When we checked on 2 October 2026, search results for "dardo studio" were led by unrelated businesses with similar names. Our structured data now carries an alternate name and a short disambiguating description, and our llms.txt names the businesses that are not us. That helps machines tell us apart, but the lasting fix is outside our site: real profiles, project write-ups and links from sites those indexes already trust. On the same day, an anonymous Bing site: search for our domain returned nothing, though nothing blocked Bingbot. Being crawlable is not the same as being indexed.

How to measure it

No single tool shows AI visibility. Combine these:

The prompt panel is our own practice, not a platform tool. Write the questions a buyer would ask, in each language you sell in, and keep the list fixed. Each month, run them on the same platforms and record the date, platform, whether you were cited, which URL, and which competitors appeared. Treat it as a dated observation, not a rank.

What no one can promise

Google says that meeting every requirement doesn't guarantee crawling, indexing or serving, and that no third-party tool has access to its internal ranking or AI systems. Bing calls Citation Share observational. Any service that promises citations in ChatGPT or AI Overviews is promising something the platforms themselves don't.

A short checklist

1. Verify your site in Search Console and confirm the Search generative AI setting is included.
2. Allow Googlebot, OAI-SearchBot, Claude-SearchBot, PerplexityBot and the user agents in robots.txt, then test at your CDN.
3. Verify in Bing Webmaster Tools and send IndexNow pings when pages change.
4. Publish pages with first-hand knowledge: one page per real need, not one per phrasing.
5. Use clear headings, tables where comparisons help, evidence with links, and real update dates.
6. Earn mentions on sites people already trust.
7. Measure monthly with the sources above.

This is the work behind our AI search optimization service, built on the same foundations as our SEO work. If you want a review of where your site stands, tell us about it.