Structured data helps AI Search but not in the way most SEO guides are telling you. The mechanism is more nuanced than “add schema and get cited,” and the landscape just shifted significantly with Google’s May 2026 deprecation of FAQ rich results. If you are relying on outdated structured data advice including advice that appeared on this site in earlier articles this update clarifies what actually works now.
The honest answer requires making three distinctions that most articles blur together: schema that creates rich results in Google Search, schema that helps AI systems understand your content, and visible content structure that drives AI citation regardless of JSON-LD markup. These are different things. They have different values. And they require different actions. Keeping up with Google’s structured data changes and want to discuss strategy with practitioners? Our community tracks every update and shares practical responses join us here.
What Changed in May 2026: The FAQ Rich Result Deprecation
On May 7, 2026, Google added a deprecation notice to its official FAQ structured data documentation at developers.google.com: “As of May 7, 2026, FAQ rich results are no longer appearing in Google Search. We will be dropping the FAQ search appearance, rich result report, and support in the Rich results test in June 2026. To allow time for adjusting your API calls, support for the FAQ rich result in the Search Console API will be removed in August 2026.”
This is a significant announcement but the industry reaction has been largely split between two overcorrections that both miss the point:
Overcorrection 1: “Schema is dead.” FAQ schema has been the foundational structured data recommendation in virtually every SEO and AI search guide published since 2021. Its deprecation feels like a repudiation of that advice. It is not. The rich result display feature is deprecated. The schema type itself is not.
Overcorrection 2: “FAQ schema is now more important than ever for AI.” Some practitioners are arguing that since the rich result is gone, the “true” value of FAQ schema for AI comprehension is revealed. The December 2024 Study from Search/Atlas found no correlation between schema markup coverage and citation rates across AI platforms. Visible Q&A formatting on the page does more measurable work than the JSON-LD layer alone.
The truth is narrower and more actionable: a display feature in Google’s traditional SERP is gone. The schema type remains valid. The more important question is whether any schema type drives AI visibility and the evidence there is more complex than most articles acknowledge.
The Three-Layer Distinction You Need to Understand
Before addressing which schema types to keep, update, or remove, it is essential to separate three layers of structured data value that are routinely conflated:
Layer 1: Schema That Creates Rich Results in Google Search
This layer is about SERP display features the visual enhancements that appear in traditional Google Search results (star ratings, FAQ dropdowns, recipe cards, event listings, etc.). These require both valid schema markup AND active Google support for that schema type as a rich result.
Google’s support for rich results has been narrowing consistently. HowTo rich results were deprecated in August 2023. FAQ rich results were deprecated in May 2026. Both followed the same pattern: widespread misuse of schema as a SERP-enhancement tool rather than a content description tool prompted Google to remove the display incentive.
Currently active Google Search rich result types (as of June 2026):
| Schema Type | Rich Result Feature | Status |
|---|---|---|
| Product + Review/AggregateRating | Product star ratings | ✅ Active |
| Article | Eligible for Top Stories | ✅ Active |
| Recipe | Recipe cards | ✅ Active |
| Video | Video results with thumbnail | ✅ Active |
| Event | Event listings | ✅ Active |
| Organization | Knowledge Panel | ✅ Active |
| LocalBusiness | Local Pack | ✅ Active |
| BreadcrumbList | Breadcrumb navigation | ✅ Active |
| FAQPage | FAQ dropdowns | ❌ Deprecated May 7, 2026 |
| HowTo | Step-by-step rich results | ❌ Deprecated Aug 2023 |
Layer 2: Schema That Helps AI Systems Understand Your Content
This layer is independent of whether Google shows a rich result for the schema type. Schema markup provides machine-readable signals about your content’s structure, authorship, dates, and entity relationships signals that both Googlebot and AI systems use during content evaluation, even when no visual SERP feature is produced.
Here is the critical nuance that the “FAQ schema is dead” camp misses: Google’s documentation still notes that FAQ structured data can stay in place, and that structured data that is not being used does not cause problems for Search. FAQPage is still a valid schema.org type. It still marks up content as structured Q&A. Whether AI systems use it differently from the JSON-LD that generates rich results is a separate question.
Layer 3: Visible Content Structure That Drives AI Citation
This is the layer that the research evidence consistently points to as the most impactful for AI citation. A December 2024 study from Search/Atlas found no correlation between schema markup coverage and citation rates across AI platforms. Visible Q&A formatting on the page does more measurable work than the JSON-LD layer alone.
In other words: writing direct, clearly structured answers to questions in your actual page content is more reliably associated with AI citations than the presence or absence of FAQPage JSON-LD. The schema may help machine interpretation at the margins, but the visible content structure is the dominant signal.
This finding aligns with the answer-first content formatting principle at the core of Hybrid Engine Optimization the framework for optimizing simultaneously for traditional search rankings and AI citation. The structured data layer adds signal; the content quality layer is the primary driver.
What Google’s Official Guidance Actually Says About Structured Data and AI
Google’s May 2026 AI optimization guide the same one that addressed LLMS.txt provides clear guidance on structured data’s role in AI Search specifically.
What the guide recommends:
- Standard schema types (Article, Organization, Product, etc.) remain recommended as part of overall technical SEO
- Entity clarity making it clear who you are, what you cover, and how your brand relates to topics is specifically called out as important for AI features
- Quality content that demonstrates expertise is the primary driver, not schema markup specifically
What the guide does NOT recommend:
- Special “AI-specific” schema types that do not exist in the official schema.org vocabulary
- Schema markup as a workaround for poor content quality
- Any specific schema type as a shortcut to AI Overview inclusion
The guide’s consistent message: AI Overviews and Google’s generative AI features use the same index and quality signals as traditional search ranking. There is no separate technical pathway for AI visibility through structured data that bypasses content quality evaluation.
This connects to how AI is fundamentally changing SEO strategy not by creating new technical shortcuts, but by raising the stakes on content quality signals that have always mattered.
Why Google Deprecated FAQ Rich Results: The Deeper Pattern
Understanding why the FAQ rich result was deprecated helps predict what might happen next and what makes a structured data investment durable.
The pattern is consistent across every recent Google schema deprecation:
HowTo rich results (deprecated August 2023): Widespread abuse. Brands adding HowTo schema to pages where the “steps” were marketing copy rather than genuine how-to instructions. The format became a SERP real estate grab rather than a user benefit.
FAQ rich results (deprecated May 2026): The same pattern. FAQ sections added to product pages, service pages, and blog posts specifically to expand SERP footprint — not because the pages genuinely answered meaningful user questions. Many FAQ sections were padded with questions no user had asked, written to trigger the dropdown display.
Google has consistently removed structured data display features that became targets for manipulation. The lesson is not to stop using structured data. It is to stop treating display features as the primary reason to implement it. Schema types that produce genuine content signals for Google and AI systems survive these changes. Schema added purely to grab SERP real estate does not.
The deprecation is a correction of an incentive problem — not a signal that structured data is declining in importance. In fact, the schema types that remain active are precisely those where the markup describes genuine content attributes (a product’s actual price and rating, a real recipe’s cooking time, an actual event’s date and location) rather than editorial constructs that could be manipulated.
What to Do With Your Existing FAQ Schema Right Now
Given the deprecation timeline, here is the specific action guidance by schema type:
FAQPage Schema: Keep or Remove?
Keep it. FAQPage as a Schema.org type is still valid, and the markup can stay on your pages without affecting search visibility.
The FAQ rich result display feature is gone. The schema itself is not harmful and may still contribute to AI system comprehension of your Q&A content. Removing it provides no benefit and creates unnecessary maintenance work.
However: Update your reporting infrastructure. Remove FAQ rich result tracking from your dashboards and auditing processes — since June 2026, Search Console no longer reports it. If your team has automated reports based on FAQ schema performance, update them before August 2026 when the Search Console API support is removed.
Important nuance: If your FAQ sections contain genuinely useful answers to questions users actually have, keep both the content and the schema. If your FAQ sections were written primarily to trigger the rich result dropdown — padded questions, marketing-focused answers, content that serves no real user need — this is an opportunity to either improve them or remove the section. The deprecation removes the SERP incentive for hollow FAQ content. Genuine FAQ content remains valuable for AI citation.
HowTo Schema: Already Deprecated — Treat Like FAQPage
HowTo rich results were deprecated in August 2023. As of May 2026, there is no HowTo rich result on any surface. Like FAQPage, HowTo is valid schema.org but produces zero Google SERP lift.
The same guidance applies: keep HowTo schema if your content genuinely describes step-by-step processes and the markup accurately reflects visible page content. Remove if the schema was added purely for the (now-gone) rich result.
What Schema Types to Prioritize Now
If you are doing a schema audit and reallocating implementation effort, prioritize these in order of confirmed value:
Tier 1 — Highest confirmed impact:
Article schema — Remains essential for editorial content. Establishes content type, publication date, modification date, and authorship. The dateModified property is particularly important: AI systems have recency preferences, and updated timestamps with corresponding content changes signal actively maintained content.
Organization schema — Establishes your brand as a recognized entity with consistent identity. The sameAs property connecting your site to your LinkedIn, Wikipedia, and social profiles is one of the strongest entity signals available. This directly affects AI systems’ ability to recognize and cite your brand as a trustworthy source.
Author/Person schema — E-E-A-T signals matter more in 2026’s AI search environment than they did in 2022. Named authors with credentials, professional profiles, and consistent publishing history are signals AI systems use to evaluate source trustworthiness. Every content piece should have Author schema with sameAs links to the author’s professional profiles.
Product + Review schema — For e-commerce, these remain among the highest-value schema implementations with active rich result support. Accurate pricing, availability, and review data are signals both Google Search and AI agents evaluating product purchases use.
Tier 2 — Valuable with active rich results:
VideoObject schema — Video content is increasingly important in AI Search. VideoObject schema with hasPart and nested Clip objects enables Key Moments display in search results — particularly valuable as Ask YouTube and Google’s video understanding capabilities expand.
LocalBusiness schema — For businesses with physical locations, LocalBusiness schema is fundamental to local search visibility and Google’s ability to connect your web presence to your Google Business Profile entity.
BreadcrumbList schema — Improves site structure understanding for both crawlers and AI systems. Low cost to implement, consistently supported.
Recipe schema — Remains an active, abuse-resistant rich result type because accurate recipe data (cooking times, ingredients, calories) is objectively verifiable and genuinely useful.
Tier 3 — Keep if already implemented, lower priority if not:
FAQPage schema — Valid, potentially helpful for AI comprehension, no longer produces rich results. Maintain existing implementations. Low priority for new implementations unless your site has substantial genuine Q&A content.
HowTo schema — Same guidance as FAQPage. Valid, no longer produces rich results. Maintain for content comprehension value.
The Real Question: What Actually Drives AI Citations?
The structured data discussion becomes more grounded when you look at what the actual data shows about AI citation behavior.
A December 2024 study from Search/Atlas found no correlation between schema markup coverage and citation rates across AI platforms. Visible Q&A formatting on the page does more measurable work than the JSON-LD layer alone.
This finding does not mean schema is useless. It means the expectation that adding more schema types directly and proportionally increases AI citation rate is not supported by evidence. What the evidence does support is that:
Content that directly answers questions in clear, structured prose gets cited more frequently. The answer-first formatting principle stating the direct answer in the first one or two sentences of each section — is more reliably correlated with AI citation than the presence of FAQPage JSON-LD.
Entity recognition drives consistent citation. Brands with strong entity signals consistent Organization schema, named authors with credentials, external brand mentions, Wikipedia presence — are more reliably cited across multiple AI platforms than brands with comprehensive schema but weak entity identity.
Topical authority generates multi-query citation. Sites with comprehensive coverage of a topic area across interconnected content clusters appear more frequently across a wider range of related queries in AI responses — not because of schema, but because of content depth. Building topical authority is the structural content investment that compounds into AI citation at scale.
Content freshness with updated dateModified supports recency signals. AI systems have recency preferences. Updating dateModified in Article schema when making substantive content changes not as a manipulation tactic, but accurately reflecting real updates contributes to how AI systems evaluate whether content is actively maintained.
A Practical Structured Data Audit for June 2026
Given everything above, here is the specific audit process for reviewing your structured data implementation right now:
Step 1: Remove FAQ rich result from all reporting and dashboards. The metric no longer exists in Google Search Console as of June 2026. Remove it from automated reports and KPI tracking to prevent confusion. Flag this for removal from Search Console API calls before August 2026.
Step 2: Audit your FAQPage schema implementations. For each page with FAQPage schema, ask: does this page have genuine Q&A content that answers questions users actually have? If yes, keep the schema. If the FAQ section was padding for the now-gone rich result, either improve the content or remove the section.
Step 3: Validate all remaining schema with Google’s Rich Results Test. Run your key pages through search.google.com/test/rich-results to catch newly deprecated properties and validate syntax. Pay particular attention to interactionStatistic — use this rather than the now-deprecated interactionCount for view counts.
Step 4: Prioritize Article schema with accurate dateModified across all editorial content. If your Article schema is missing dateModified or using a static date that never updates, fix this. Accurate modification timestamps contribute to recency signals that AI systems evaluate.
Step 5: Audit Organization schema for completeness. Ensure your Organization schema includes current sameAs links to all active professional profiles — LinkedIn, Wikipedia (if applicable), relevant industry directories, and social profiles. These entity links are how AI systems connect your on-site claims to verified external identity.
Step 6: Review Author schema across your content library. Named authors with complete sameAs links to their professional profiles are an E-E-A-T signal that directly influences AI citation authority. If content is attributed to an author with a bare name and no professional profile links, update it.
The Broader Strategic Lesson From the FAQ Deprecation
The FAQ rich result deprecation is the second act of a pattern that started with HowTo in 2023. Both schema types were deprecated because widespread abuse corrupted their signal value — teams adding the markup primarily for SERP real estate rather than to accurately describe genuine content.
This pattern suggests a durable principle for structured data strategy: implement schema that accurately describes real content attributes, not schema designed to trigger display features. The schema types Google has deprecated are precisely those where the incentive structure pointed toward manipulation. The schema types that remain active are those where the markup is objectively verifiable against real content.
For AI search specifically, the same principle applies with even more force. AI systems evaluate content comprehensively — they can read your visible page structure, your prose quality, and your entity signals independently of your JSON-LD. Schema that accurately describes well-structured content adds a useful signal. Schema applied as a technical trick on top of poor content adds confusion.
The structured data principles that help AI systems understand your content remain valid — what has changed is the expectation that specific schema types translate directly into SERP display features or guaranteed AI citations. The direct path to AI visibility runs through content quality and authority, with structured data as a supporting signal.
FAQs
Is FAQ schema dead after the May 2026 deprecation?
No. Only the FAQ rich result display feature in Google Search is deprecated. FAQPage remains a valid schema.org type and can be kept on your pages without any negative SEO impact. The structured data itself may still help AI systems understand your Q&A content, even without a visual rich result in traditional Google Search.
Should I remove FAQPage schema from all my pages?
Not necessarily. Google has confirmed that unused structured data does not cause problems for Search. If your FAQ content is genuinely useful and accurately described by the schema, keep both. If your FAQ sections were created primarily to trigger the now-deprecated rich result dropdown, evaluate whether the content has genuine value independent of that display feature.
Does FAQ schema still help with Google AI Overviews?
There is no confirmed direct correlation between FAQPage schema specifically and AI Overview inclusion. Research suggests visible, well-structured Q&A formatting in your page content is a stronger predictor of AI citation than the presence of JSON-LD FAQPage markup. The content quality and topical authority signals that drive traditional rankings are the primary determinants of AI Overview eligibility.
Which schema types are currently most valuable for AI search visibility?
Article schema with accurate dateModified, Organization schema with comprehensive sameAs entity links, and Author schema with professional profile connections are the highest-impact schema investments for AI search visibility specifically because they strengthen entity recognition and content freshness signals. Product schema with reviews remains the top priority for e-commerce.
What happened to HowTo schema?
HowTo rich results were deprecated in August 2023. As of June 2026, HowTo schema still exists as a valid schema.org type but produces no rich results in Google Search. Keep existing HowTo implementations for content comprehension value; it is not a priority for new implementation.
Does adding more schema types improve AI citation rates?
The evidence does not support this. Research shows no correlation between schema markup coverage and AI citation rates across platforms. Visible, well-structured content answers, strong topical authority, and entity recognition signals are better predictors of AI citation frequency than schema coverage volume.
What should I do about Search Console reporting for FAQ schema?
Remove FAQ rich result tracking from your dashboards and reporting processes. The Search Console FAQ rich result report is being removed in June 2026. If you have automated API calls pulling FAQ rich result data, update those calls before August 2026 when the Search Console API support is removed.
Are there any new schema types specifically for AI search I should add?
No. Google’s May 2026 optimization guide groups “special schema for AI” alongside LLMS.txt as tactics that are not needed and not recommended. There are no officially recognized schema types designed specifically for AI search. The standard schema vocabulary applied accurately to genuine content is what works.
Conclusion
The FAQ rich result deprecation is a correction, not a collapse. Google removed a display feature that had been widely abused as a SERP enhancement tactic. The underlying value of structured data for machine comprehension helping Google and AI systems understand what your content is about, who wrote it, when it was updated, and how it relates to the entities in your space remains.
The practical response to May 2026 is straightforward: update your reporting to reflect the removed FAQ rich result, audit FAQPage implementations for genuine content value, and prioritize the schema types that produce both active rich results and strong entity comprehension signals Article, Organization, Author, Product, and Video.
More importantly: do not let schema chasing displace the work that actually drives AI visibility. The current best practice clear prose, direct answers, structured headings, and entity clarity — will likely outlast any specific schema-type advice. The brands that earn consistent AI citations in 2026 and beyond are those building genuine content authority, not those with the most comprehensive JSON-LD implementation.
Understanding how AI search is transforming what SEO actually requires provides the strategic context that makes each technical change including the FAQ deprecation interpretable rather than alarming. Individual tactics evolve. The principle that quality content with clear structure, expert attribution, and genuine topical depth earns visibility does not. Auditing your structured data in response to the FAQ deprecation and want expert input on priorities? Our community includes technical SEOs actively navigating these changes join us here.
