No, not in any way a controlled test has been able to detect so far. Schema markup, the structured data code added to a webpage, governs whether that page is eligible for a rich result in classic Google search, not whether an AI model decides to cite it. Google has said directly that structured data does not improve ranking, and the one large controlled experiment built specifically to test its effect on AI citations found no measurable gain in Google's AI Mode or in ChatGPT, and a small decline in Google AI Overviews. If someone is selling you a special schema format as the key to appearing in AI answers, the evidence available right now does not support that claim.
What is structured data actually supposed to do?
Help machines understand a page's content well enough to display it as a rich result, nothing more than that by Google's own description.
Google's own developer documentation on structured data says Google uses the markup it finds to understand the content of a page, and that adding it "can enable search results that are more engaging to users." The documentation ties the benefit specifically to eligibility for enhanced display, visual extras like star ratings, event dates, or recipe cards in a normal search result, and it warns against describing content that is not actually visible on the page, which implies the markup does not make content appear anywhere it was not already shown. Ranking is never mentioned in that guidance. Neither is AI Overviews or AI Mode.
Has Google ever said schema affects ranking either way?
Yes, repeatedly, and the answer is no.
John Mueller, who coordinates Google's Search Relations team, posted on Bluesky in April 2025 that "structured data won't make your site rank better," adding that using it for other purposes "won't cause problems, but you're unlikely to see any visible change from it in Google Search." That matches guidance Google has given since at least 2018, and it is consistent with what the developer documentation above actually says: the benefit is rich-result eligibility and machine comprehension of entities on a page, not a ranking boost.
Source: Ahrefs, "Does Schema Help You Get Cited by AI?" (May 11, 2026), linked in the article.
Is there any direct evidence on AI citations specifically?
One study, and it is worth reading carefully rather than just the headline.
Ahrefs published a controlled experiment on schema and AI citations in May 2026. The team tracked 1,885 pages that added JSON-LD structured data between August 2025 and March 2026, and compared them against matched control pages using a difference-in-differences method, the same statistical approach researchers use to isolate one change's effect from everything else happening at the same time. The result: citations in Google's AI Mode moved +2.4% and in ChatGPT +2.2%, both changes small enough that Ahrefs describes them as statistically indistinguishable from zero. Citations in Google AI Overviews actually fell 4.6%, a decline the authors call small but statistically significant, working out to an average loss of around 12 daily citations per page in that group.
Ahrefs is upfront about the limits of its own study. Every page in the sample already had 100 or more AI Overview citations before the schema was added, so the test cannot say anything about whether schema helps a page that has not been discovered by these systems yet. Pages that add schema often change other things on the page at the same time, which the study could not fully separate out, and the team pooled every type of schema together rather than testing specific types individually. Their own conclusion: "we can't tell whether the schema did a tiny bit of good or nothing at all." That is a more honest statement than most marketing claims about AI visibility offer, and it is the most direct evidence that exists on this question right now.
So why do AI-cited pages so often have schema markup on them?
Because the kind of page that earns citations in the first place also tends to be the kind of page a careful publisher marks up, which is a correlation, not a cause.
Ahrefs' earlier analysis of six million URLs found that pages already cited by AI systems were far more likely to carry schema markup than pages that were not. That looks like proof schema works, until you remember that well-built, well-maintained, authoritative pages tend to do several things well at once: clear writing, accurate facts, good technical hygiene, and yes, structured data. The controlled experiment above is what separates correlation from cause, by watching what happened to citations after schema was added to pages that did not already have it, and the answer was close to nothing. The same caution applies broadly to how AI assistants pick what to cite: the visible technical signal is rarely the actual reason a source gets used.
What should you actually do with your development time instead?
Spend it on the things that have real evidence behind them: being written about by people who do not work for you, and making sure a crawler can actually read your site at all.
Add schema where it buys you a legitimate rich result, a product listing, an event, a FAQ block people will actually read, because that is a real, documented benefit with a clear mechanism. Do not add it expecting it to move your AI citations, and do not pay anyone a premium for a "special AI schema," because the structured data vocabulary available to every site is the same one Google documents publicly at schema.org, and Google's own guidance for AI features says there is no special markup or file needed to appear in them. What actually drives citations, across the evidence this site has reviewed, is independent coverage of your company, a clean and crawlable site, and a track record other sources repeat. That is the same conclusion behind why earned media carries so much weight in AI answers, and it holds whether or not a single line of JSON-LD sits in your page's head tag. If you want a sense of how a page without heavy markup can still show up in Google AI Overviews or get picked up by an llms.txt file worth building for other reasons, both of those questions have real answers too, just not the one schema markup promises.
The direct answer, for anyone who read this far
Structured data makes your page eligible for rich results in classic Google search and helps machines parse what is on the page. It is not a proven lever for AI citations. Google's own staff say it does not affect ranking, and the one controlled study built to test its effect on AI Mode and ChatGPT citations found no measurable gain, with a small measured decline in AI Overviews. Build schema for the rich results it is actually designed to produce, and put your real effort into earned coverage and a crawlable site, which is what the evidence currently supports. If you want an honest read on what is actually driving your AI visibility right now, talk to us.
