Google Search Console Generative AI Performance Reports: A Complete Guide (2026)

Google Search Console now has a generative AI report

On June 3, 2026, Google added a new report section to Search Console.^[1] Search Generative AI Performance Reports — the first official data showing how often your site surfaces in generative AI features inside AI Overviews, AI Mode, and Discover. The gap that had forced AEO practitioners to estimate "AI traffic" with third-party tools is now filled directly by Google.

Why does this report matter? As AI search spreads, "is my content cited in AI answers" has become a primary visibility metric. But the existing Search Console performance report did not separate generative AI features from regular organic search. The new report fills that void — though the limits of its early version are clear. This piece lays out what the report does and does not show, and how to connect it to AEO strategy.

---

A 30-second definition — key terms

Search Generative AI Performance Report is a dedicated section inside Google Search Console — a measurement tool that shows how many times your site's URLs were exposed (impressions) in generative AI features inside AI Overviews, AI Mode, and Discover, along with their dimensions (page, country, device, date).^[1]

AI Overviews is the feature that shows an AI-generated summary answer at the top of Google search results, citing and referencing source URLs.

AI Mode is Google's generative AI search interface for searching through conversational question-and-answer.

Generative AI opt-out toggle is a control that lets site owners selectively block exposure in generative AI features inside AI Overviews, AI Mode, and Discover from within Search Console. It took practical effect on June 17, 2026.^[4]

---

The report structure at a glance

Source: Google Search Central Blog (2026.06.03)

---

Provided dimensions vs. dimensions not provided

You have to grasp what the report gives and what it withholds before you can build a measurement strategy.^[2][3]

The absence of clicks and queries is the core constraint. You can know the fact that "my page was exposed in an AI answer," but you cannot confirm which question produced that exposure or whether it led to an actual click.^[5] To fill this gap, you need to run third-party AEO measurement tools in parallel.

---

The backdrop — the UK CMA's binding requirement

This report's launch is not a pure Google-initiated move. It is the result of a binding measure the UK's Competition and Markets Authority (CMA) imposed on Google under the UK Digital Markets, Competition and Consumers Act (DMCCA) 2024.^[4] The CMA officially described it as "the first case in the world enforcing a mandatory opt-out right for AI-based search." Google must fully implement the related requirements by around March 2027.

The initial release was limited to a small set of UK site owners,^[1] and the timeline for a global expansion has not been officially confirmed. That said, given the CMA requirement in the background, a global rollout looks like a matter of time.

---

The opt-out toggle — blocking scope and limits

The generative AI opt-out toggle, launched the same day, gives site owners a choice over AI exposure. It took practical effect on June 17, 2026,^[4] and the key points are as follows.

The important point is that the opt-out toggle does not apply to the Gemini app.^[4] In other words, even if you block your content in AI Overviews, your content can still surface in Gemini conversations. For publishers, this means having to manage "Google Search AI" and "Gemini AI" as separate channels.

---

How to connect this to AEO strategy

Use impressions as a baseline

Impression data is not a traffic metric but a content resonance signal. Exposure happens when AI judges your content valuable enough to cite. So pages with many impressions = content that AI classifies as high quality, and that list becomes a priority map for content improvement.

A practical approach looks like this:

1. As soon as you can access the AI report, export the 90-day impression data and set a baseline.
2. Cross-analyze it with the clicks for the same URLs in the standard Search performance report to estimate whether AI exposure leads to real traffic contribution.
3. Review the distribution of AI exposure by country to set domestic vs. international AEO priorities.

Content structure that raises AI citations

Google's official AI optimization guide, published in May 2026, makes its core message clear: "Optimizing for AI search is optimizing for the search experience — and it's still SEO."^[6] What matters is the crawlability, structure, and trustworthiness of the content itself, ahead of tricks (content chunking, creating unnecessary llms.txt files).

Concretely, the factors that raise the likelihood of an AI citation are as follows.

Run it alongside third-party tools

The GSC generative AI report provides only impressions, so you cannot tell which query produced the exposure. For query-level analysis, you need to pair it with the third-party solutions covered in AI visibility monitoring tools compared (Semrush AI Toolkit, Ahrefs AI Overview tracking, BOIDA BVI, and others). BOIDA is a solution that tracks multi-channel AI responses across ChatGPT, Claude, Gemini, Perplexity, Grok, DeepSeek, and more, extending the measurement scope to non-Google AI channels that GSC data cannot cover.

---

Action steps — what you can do now

Even without report access yet, you can prepare in advance.

Step 1: Audit content structure (possible immediately)

• Check that every comparison table is a markdown table (| col | col |), not an image.
• Review whether your H2/H3 headings take the form of actual user questions ("what is ~", "how do I ~").
• Confirm whether you have an FAQ section and whether FAQ schema (FAQPage) is applied — see The complete guide to structured data for AEO.

Step 2: Prepare baseline data (as soon as you get access)

• The moment access is granted, export and archive a 90-day AI impressions CSV.
• Export the standard Search performance report (clicks, impressions) data for the same period in parallel to prepare for cross-analysis.

Step 3: Set priorities by country

• Use the country dimension data to identify markets where AI Overviews is active.
• For domestic (Korea) AI visibility, separately monitor Naver AI Briefing beyond Google — see Naver AI Briefing optimization.

Step 4: Decide on opt-out (a strategic judgment is needed)

• If you judge that generative AI exposure carries more brand risk (e.g., information distortion) than traffic, consider opting out.
• For most publishers, keeping AI exposure plus strengthening content quality is the better move.

---

Why this report changes the AEO measurement paradigm

As laid out in What is AEO, AEO aims for "having your content cited and referenced when AI generates an answer." Until now there has been no official tool to measure whether that goal is achieved. Only indirect methods existed — third-party scraping, manual query tests, GA4 referrer analysis, and the like.^[3]

The GSC generative AI report is the first official solution to that void.^[1] It is not perfect — there are no clicks and no queries. But the metric itself, "how many times my URL was exposed in Google AI features," becomes a new yardstick for measuring content resonance. As covered in The global GEO/AEO landscape 2026, in a trend where AI search is rising as a primary gateway to traffic, the data value of this report will only grow as clicks are added.

---

FAQ

Q. When did the Generative AI performance reports launch? They were officially announced on June 3, 2026 via the Google Search Central Blog, in a limited release to a small set of UK site owners.^[1]

Q. Can I see clicks data in the report? Clicks, CTR, queries, and position are not provided right now. Only impressions plus the page, country, device, and date dimensions are available.^[2]

Q. Does turning on the opt-out toggle also affect regular search ranking? No. Google has officially confirmed that the opt-out setting applies only to generative AI features such as AI Overviews, AI Mode, and Discover, and is not used as a signal for regular organic search ranking.^[4]

Q. How should I structure content to improve AI visibility? Question-based H2/H3 structure, real markdown tables, FAQ blocks, and short sentences raise how often AI cites you. Google's official AI optimization guide names content crawlability and structuring as core requirements.^[6]

Q. How do Korean AEO solutions like BOIDA use this report? They can use the impressions from the GSC generative AI report as a baseline and track how AI citations change before and after content-structure improvements. Multi-channel tracking solutions like BOIDA extend the measurement scope to non-Google AI (ChatGPT, Perplexity, etc.) that GSC does not cover.

Q. What should I do if the report isn't in my account yet? As of June 2026 it is in a limited release to a small set of UK sites. The timeline for a global expansion has not been officially confirmed, so you should periodically check the Google Search Central Blog for updates.^[1]

---

In summary

The Google Search Console Generative AI performance report is the starting point for AEO measurement. It has the constraint of no clicks and no queries, but the fact that it is the first tool to confirm "is my content exposed in AI features" through official data does not change. What you can do now is clear: the moment access arrives, secure baseline data immediately, audit your content structure (real tables, FAQ, short sentences), and put in place a system that runs alongside third-party AI measurement tools.