Schema Markup for AI Overviews: Requirements and Limits

There is no special schema type that guarantees an AI Overview citation. Choose structured data because it accurately describes the page and supports a documented use case, not because a checklist says every article needs FAQ markup.

Google's AI-feature documentation says no special schema is required for these features. Google's Search documentation changelog also records the retirement of FAQ rich results in May 2026. Old promises of FAQ expansions in Google results should not drive a new implementation.

Choose the representation from the content

Page	Possible representation to evaluate	Evidence to preserve
Original article	Article/BlogPosting	Real author, actual publication/update dates
Business homepage	Organization, or appropriate LocalBusiness	Correct identity and contact details
Product page	Applicable product markup	Visible product/offer information
A list of questions	FAQPage vocabulary if useful to your consumers	Visible questions and answers; no Google FAQ-result promise

These are starting points, not a substitute for the relevant current feature specification. A schema.org type can exist without having a Google rich-result feature.

A truthful article example

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "How to Inspect a Sitemap Response",
  "datePublished": "2026-09-28",
  "author": {"@type": "Organization", "name": "Example Publisher"}
}

This fictional block illustrates a small description of an article. It is not a complete rich-result eligibility template. Replace every field with actual page information and review current Article documentation for the intended use case.

Validate three different things

Syntax: can a JSON parser read the block?
Meaning: do the type and fields match visible content and real entities?
Feature eligibility: does the page meet the current requirements of the specific consumer/search feature?

A schema validator passing the first two does not prove a rich result will appear, let alone an AI citation. Inspect the published page after a theme/plugin change to catch duplicated or stale blocks.

The schema generator helps prepare supported markup. Check its output against your actual content and the current documentation; it does not certify AI-search eligibility or results.

Measure the reader improvement separately

If your page lacks a useful example, add that example because it completes the reader's task. Do not attribute later citation changes to the markup without evidence that separates other causes. Use the citation observation workflow and keep a dated record of the content and markup changes.

For implementation, use schema validation diagnosis. For planning, use the AI-search content checklist.