FAQ Schema Is Dead in Google Search — Here's What Actually Gets You Cited by AI in 2026

Short answer: Google confirmed on May 7, 2026 that FAQ rich results no longer appear in Search, with Rich Results Test support removed in June and API support dropped in August. But that's a Google-Search-visibility change, not an AI-citation change. FAQ schema still helps Bing, Perplexity, ChatGPT Search and Google's own AI systems parse and quote your answers — and the schema types that actually earn AI citations look very different from the old rich-result playbook.

What Actually Changed on May 7, 2026

Google Search Central's announcement was narrow and specific:

• FAQ rich results (the accordion snippets under blue links) no longer render in Search.
• The Rich Result Test dropped FAQPage support in June 2026.
• Search Console's FAQ rich-result report was retired.
• API support was fully removed in August 2026.

What did not change: Google still parses FAQPage structured data to understand page content. Bing, Perplexity and ChatGPT Search continue to use it for extraction and citation. Voice assistants use it for Speakable pairings. The markup is still valuable — it just no longer earns you visual real estate in the SERP.

Why Publishers Are Overreacting

Every 6 months a schema type gets retired and half the SEO industry declares "schema is dead." It never is. What dies is the visual rich result. The parsing benefit persists — and in an era where AI engines quote extractable Q&A blocks at ~2.5× the rate of prose paragraphs, the markup that identifies "this is a question and this is its answer" is more valuable than ever.

The Schema Stack That Actually Earns AI Citations in 2026

1. Article / BlogPosting — with the fields AI actually reads

Not just the type. AI engines heavily weight author (with @type: Person and a sameAs to a real profile), datePublished, and dateModified. Freshness is a citation signal — Perplexity has a strong recency bias and ChatGPT Search boosts recent content on time-sensitive queries.

2. Organization — with sameAs pointing to Wikidata and LinkedIn

This is the single most under-used AI-citation signal. When your Organization schema includes:

"sameAs": [
  "https://www.wikidata.org/wiki/Q...",
  "https://www.linkedin.com/company/...",
  "https://www.crunchbase.com/organization/..."
]

AI engines link your website to your entity in their knowledge graph. Without it, you're an anonymous domain. With it, you're a named source they can attribute — which is a precondition for being cited by name.

3. FAQPage — kept, not killed

Keep FAQPage on pages where you have genuine Q&A. Answer in 40–60 words per question — that's the extraction window AI engines quote from most reliably.

4. HowTo — for procedural queries

Any "how to" content should include HowTo schema. How-to queries trigger AI Mode at the highest rate of any query type, and HowTo schema makes each step individually citable.

5. Speakable — for voice

Nominates specific sections as spoken-answer candidates for Google Assistant, Siri (via Applebot-Extended), and Alexa. Low cost, real upside as voice AI grows.

What to Do This Week

1. Do not remove FAQPage schema. It still helps AI parsing.
2. Audit Organization schema. Add sameAs links to Wikidata, LinkedIn, Crunchbase, and any other entity graphs where your business is listed.
3. Rewrite FAQ answers to 40–60 words. That's the extraction sweet spot for AI quoting.
4. Add HowTo schema to procedural pages. High leverage for AI Mode visibility.
5. Add author schema with real sameAs. Ghost-authored posts don't earn E-E-A-T signals.

The Broader Point

The old rich-result game rewarded visual real estate in Google's SERP. The new AI-citation game rewards being parseable, attributable, and fresh. Same underlying signals (schema, author, dates, entity links) — different payoff surface. Publishers still treating schema as "the rich result trick" are behind. Publishers treating it as "how I tell AI engines who I am and what I said" are winning.