Schema Markup for AI Search: What Actually Helps and What's Myth

Schema markup for AI search is not a secret switch: Google states plainly that there is no special structured data you need to add to appear in AI Overviews or AI Mode. That doesn't make schema useless. It still describes who wrote a page, which organization runs the site and how pages fit together, and that context helps search systems read you correctly. This guide separates what the official docs say from what agency blogs claim, and gives a small blog a markup setup it can finish in an afternoon.

Published: 2026-10-16 · 7 min read · Satelo

What Google actually says about schema and AI features

Google's page AI features and your website is the primary source here, and it is unusually direct. It says there are no additional requirements to appear in AI Overviews or AI Mode, and that there is no special schema.org structured data you need to add. To be eligible as a supporting link, a page must be indexed and eligible to be shown in Google Search with a snippet. That's it.

The same page lists ordinary SEO best practices that still apply, and one of them is about markup: make sure your structured data matches the visible text on the page. So Google's position has two halves:

• Schema is not a ticket into AI answers. No markup type makes a page "AI eligible."
• If you use schema, it has to be honest. Markup that disagrees with what readers see is a problem, not a boost.

Anyone who tells you that adding a certain JSON-LD block will "get you into AI Overviews" is contradicting Google's own documentation.

So why bother with structured data at all?

Google's introduction to structured data describes it as a standardized format for providing information about a page and classifying its content. In plain terms, it removes guesswork. Without it, a crawler has to infer that "Jane Kim" at the top of a post is the author and not a person the post is about. With it, you say so explicitly.

That has three practical benefits for a blog:

1. Entity clarity. Organization markup tells Google which name, logo and profiles belong to your site. Google's Organization documentation says it can help Google understand your organization's administrative details and disambiguate it in search results.
2. Accurate facts. Article markup can state the author, publish date and last-modified date in a form machines don't have to parse out of your design.
3. Rich result eligibility. Some markup types still unlock visual features in classic search results, like breadcrumbs. Eligibility is not a guarantee, as we'll see below.

None of this is AI-specific. It is the same groundwork that makes any search system, classic or generative, less likely to get basic facts about you wrong.

The schema a blog actually needs

Most small blogs need four types. Anything beyond that should be justified by the content of a specific page.

Two details from the docs are worth knowing. First, Google's Article page says there are no required properties; you add the ones that apply to your content, and Article objects must be based on Article, NewsArticle or BlogPosting. Second, the Organization page recommends placing that markup on your home page or a single page that describes your organization, and says you don't need it on every page.

Google also recommends JSON-LD as the format when your setup allows it. A minimal BlogPosting block, for illustration, looks like this:

{"@context": "https://schema.org", "@type": "BlogPosting", "headline": "How We Grow Tomatoes on a Balcony", "datePublished": "2026-10-01", "dateModified": "2026-10-12", "author": {"@type": "Person", "name": "Jane Kim", "url": "https://example.com/about/jane/"}}

Every value in it must be visible on the page: the same headline, the same author byline, the same dates.

The visible-content rule (where most "AI schema" advice breaks)

Google's general structured data guidelines say: don't mark up content that is not visible to readers of the page, and your structured data must be a true representation of the page content. The intro page adds that you shouldn't mark up information that isn't visible to the user, even if it's accurate.

That rules out several tactics that circulate as "AI optimization":

• Stuffing a hidden FAQPage block with questions that don't appear on the page.
• Adding author credentials, awards or ratings in markup that the page itself never mentions.
• Marking up reviews or aggregate ratings you collected nowhere.
• Copying the same "about" text into markup on every post.

The guidelines are explicit that a structured data issue can result in a manual action. Per the same page, such an action removes rich result eligibility but doesn't affect how the page ranks in web search. Still, it means you did work that bought you a penalty notice instead of anything useful.

Rich results are shrinking, and that matters for this debate

A lot of older schema advice is built around FAQ snippets. That advice is now out of date. Google's Search Central changelog shows that in May 2026 Google deprecated the FAQ rich result, which stopped appearing in Google Search starting May 7, 2026, and the documentation for it was later removed.

Two takeaways:

• An FAQ section is still good for readers, and you can keep it on the page. Just don't expect FAQPage markup to win you extra space in Google.
• Even where a rich result exists, Google's guidelines say it does not guarantee your structured data will show up in results; correct markup enables a feature, it doesn't promise it.

When someone claims a schema type "increases AI citations by X%," ask for the source. We could not find any official documentation from Google, OpenAI or Microsoft that publishes such a figure.

What about ChatGPT, Copilot and other AI systems?

The honest answer: there is much less public documentation than for Google, so be skeptical of confident claims.

• ChatGPT search. OpenAI's crawler documentation says OAI-SearchBot is used to surface websites in ChatGPT's search features and recommends allowing it in robots.txt. It does not mention structured data at all. Crawl access comes first; see our guide on getting your site indexed by ChatGPT.
• Microsoft Copilot. Copilot draws on Bing, and Bing has its own webmaster guidelines and markup tools. We did not find an official Microsoft statement that schema raises Copilot citations; our Copilot guide covers what Microsoft does document.
• Other assistants. Treat any "schema for Perplexity" or "schema for Claude" checklist as a hypothesis unless it links to the vendor's own docs.

What every system does need is readable HTML with the answer in the text. Markup can label that text; it can't replace it. The same applies to newer files like llms.txt, which we discuss in the llms.txt guide.

Myths vs. what the docs support

How to add and check schema on a small blog

1. See what you already have. Many themes and SEO plugins output Article and Organization markup automatically. Run a post through Google's Rich Results Test before adding anything, so you don't end up with duplicate blocks.
2. Fix the Organization block once. Correct name, logo URL and real profile links, on the home page or About page.
3. Make the author real. Every post should show a byline that links to a bio page, and the Person markup should point to that same URL.
4. Keep dates honest. Update dateModified only when the content meaningfully changed, and show that date on the page.
5. Validate. Use the Rich Results Test for Google features and the Schema Markup Validator for general schema.org syntax.
6. Watch Search Console. Google's intro page recommends the rich result status reports after you deploy; they flag errors across your whole site.

Then spend the rest of your time on what AI answers actually quote: clear, sourced paragraphs that answer a question directly. We cover that in how to get cited by ChatGPT and in AI Overviews and brand mentions, since being mentioned on other relevant sites is part of how AI systems learn who you are.

What to do next

Run one post through the Rich Results Test today and fix whatever doesn't match the visible page. Then work on the part schema can't do for you: being talked about on other sites in your niche. If you run WordPress, the Satelo plugin connects your blog in one click to a small public-beta network where related sites publish partner articles about each other in a circle. Articles are labeled as partner content with rel="sponsored" links, you can review drafts before they go live, and each one you host earns satellites to spend on articles about your own blog.

FAQ

Does schema markup help with AI search?

Not as a direct ticket in. Google says no special schema is needed to appear in AI Overviews or AI Mode. Schema still helps search systems identify your authors, organization and page structure, which supports accurate understanding of your content.

What schema should I use for AI Overviews?

None specifically. Google's AI features documentation says there are no additional requirements beyond being indexed and eligible for a snippet. Use normal types that fit your content, such as Article, Organization and BreadcrumbList.

Does ChatGPT use schema markup?

OpenAI's crawler documentation doesn't mention structured data. It recommends allowing OAI-SearchBot in robots.txt so your site can appear in ChatGPT search. Make sure your content is crawlable and readable as plain HTML first.

Is FAQ schema still worth adding?

Not for Google rich results: the FAQ rich result stopped appearing in Google Search in May 2026. A visible FAQ section can still help readers, and the markup is harmless if it matches the page, but don't expect extra search space from it.

Can wrong schema markup hurt my site?

Markup that misrepresents the page can lead to a manual action. According to Google's guidelines, that removes rich result eligibility but doesn't affect web search ranking. Keep markup identical to what readers see.

How do I test my structured data?

Use Google's Rich Results Test for Google-supported features and the Schema Markup Validator for general schema.org syntax. After deployment, check the rich result reports in Search Console for site-wide errors.

Read next

Nofollow vs sponsored vs UGC: which rel attribute to useNofollow vs sponsored vs UGC explained: what each rel attribute means, when Google expects it, how to combine them and how to set them up in WordPress.

AI Overviews brand mentions: how to get your brand into Google’s AI answersHow AI Overviews brand mentions work, what Google officially says about appearing in AI features, and a practical plan to get your brand named and linked.

llms.txt: what it is and whether your site needs oneAn honest llms.txt guide: what the proposed standard is, a ready template, how it differs from robots.txt and what it can and cannot do for AI visibility today.

How to get cited by ChatGPTHow to get cited by ChatGPT and other AI assistants: let the right crawlers in, write answer-ready pages and earn mentions on sites the models already trust.