Google confirmed on November 11, 2026, what several SEOs quickly started calling the Google AI report logging error. The bug cut reported impressions in the Generative AI performance report inside Search Console, for data covering November 9 through November 11. It is the third confirmed bug to hit that specific report since Google launched it in June, and it landed in the middle of the November 2026 core update.
What Google Confirmed About The Google AI Report Logging Error
Google posted the fix notice on its data anomalies page. That’s the same page it uses to log every known Search Console reporting problem. The entry says plainly: “A logging error caused a decrease in impressions on the Generative AI performance report in Search for data from November 9 – November 11, 2026.”
That single sentence matters more than it looks. Google also added a second line: the issue “affects data logging only.” In plain terms, your site’s actual appearances in AI Overviews and AI Mode did not change during those three days. Only the count that Search Console shows you went down.
As of this writing, Google has not posted a resolution date or confirmed whether the missing data will be backfilled. Search Engine Roundtable tracks these notices daily, and it flagged the entry shortly after it appeared. Site owners who open the report will see an annotation directly on the chart for those dates. That detail comes from a report by Search Engine Roundtable.
Google never explains the mechanics behind these errors in detail. The anomalies page exists for exactly one purpose: to tell you a number is wrong without telling you why the pipeline broke. In my experience, that’s typical for any large logging system built on top of a feature that’s still being rolled out and changed every few weeks. A dashboard like this one sits between live ranking systems and a reporting warehouse. A change on either side can quietly throw off the count without touching what actually happened in the results page.
What Google did not say is just as telling. There’s no mention of a specific percentage drop, no note about which countries or page types were hit hardest, and no estimate of how many properties noticed it. That’s consistent with how Google has handled the two earlier incidents too. It confirms a problem exists and roughly when, and leaves the scale of it to third parties to guess at from their own dashboards.
Why The Timing Made It Confusing
The bug window overlapped almost exactly with the start of the November 2026 core update, which Google began rolling out on November 10. Anyone glancing at a dip in AI impressions that week had two plausible explanations fighting for attention. One was a core update reshuffling rankings. The other was a reporting bug that had nothing to do with rankings at all.
Our November 2026 core update checklist covers what that update actually touches. Read it side by side with the anomalies page before you assume a drop is ranking-related. Search rankings were also unusually volatile that week for reasons that had nothing to do with the logging bug.
The lesson here isn’t complicated, but it is easy to skip when you’re staring at a chart that just dropped. A number going down in one report, during a week when something else big is happening, is not proof that the two are connected. You have to check both causes separately before you act on either one.
There was a third factor in the mix that week too. Google had also pushed Gemini 4 Hydrogen into wider use as the default model behind AI Overviews. We covered that change in our report on Gemini 4 Hydrogen becoming Google’s new AI Mode default. A model swap like that can genuinely shift which pages get cited. Site owners watching their AI impressions that week had three moving parts to untangle at once: a core update, a model change, and a reporting bug. Treating each one on its own, rather than blending them into one vague sense that “something changed,” is the only way to figure out which one actually touched your traffic.
What The Generative AI Performance Report Actually Tracks
Google introduced this report in a Search Central Blog post in June 2026, and it fills a gap that site owners had been asking for since AI Overviews first launched. It shows how often your pages were cited inside AI Overviews and AI Mode on Google Search, broken down by page, country, date, and device. Google described the feature in its original announcement.
A few limits matter for anyone relying on it. The report only counts impressions, meaning a page showed up inside an AI answer. It does not report clicks, click-through rate, or the actual queries that triggered the citation. Those numbers already live inside the regular Performance report, mixed in with all your other web traffic, with no clean way to pull the AI-driven clicks out on their own.
That also means the Generative AI report and the classic Performance report are not meant to be added together. The AI impressions you see in the new report are already counted inside your web totals elsewhere. Treat the new report as a magnifying glass on a slice of data you already had, not a second, separate traffic source.
Our breakdown of AI Overview coverage hitting roughly 80 percent of searches in 2026 gives useful context for why Google built a dedicated report for this in the first place. And if you’re setting up a regular habit of checking this data, our guide to tracking your brand’s presence in AI Mode walks through a simple weekly routine.
How The Rollout Got Here Before The Bug Hit
The report didn’t appear everywhere at once. Google’s June announcement described an initial release to “a subset of websites,” a common way Google tests a new report before trusting it at full scale. Over the following months, access widened in stages, with Google adding the ability to filter by multimodal, image-based searches alongside the original text-based view in a separate update in September.
That staged rollout explains part of why the bugs cluster the way they do. Each expansion brings a fresh wave of properties into the pipeline for the first time, and a pipeline handling new volume is exactly where a logging error tends to surface. The August 13 bug landed within days of the report reaching full worldwide availability, which lines up with that pattern. November’s bug didn’t follow a known expansion, which is one reason it’s harder to explain away.
If you manage a site that only recently started seeing this report appear in Search Console, it’s worth treating your first few weeks of data with a bit of caution. Early numbers from any newly rolled-out report are more likely to catch a bug before the pipeline settles down.
How To Tell A Logging Bug From A Real Visibility Drop
Most site owners don’t have the time to read Google’s anomalies page every week. Here is a shorter version of what to actually do when the Generative AI performance report dips.
First, check the anomalies page before you check anything else. It is the fastest way to rule out a known bug in under a minute, and Google updates it the same day it confirms a problem.
Second, look at the date range. A drop that lines up exactly with a dated entry on that page is almost certainly the bug, not your content. A drop that starts a week before or after a known bug entry deserves a real look.
Third, cross-check with a source outside Search Console. If your own analytics, a rank tracker, or a third-party AI visibility tool shows the same pages losing ground over the same period, that’s a real signal. Trust it more than Search Console data alone, especially during a known logging error.
Fourth, be patient about backfilled data. When Google fixed the August 13 to 17 bug, it restored the missing data about a week later and said so directly on the anomalies page. There is no guarantee November’s gap gets the same treatment, but it is the closest precedent available right now.
None of this requires a dashboard or a new tool. It just requires checking one Google page before reacting to a different one, which is a five-minute habit most teams skip under deadline pressure.
The Google AI Report Logging Error Has Happened Before
The November incident is not an isolated event. It is the third confirmed logging error tied to this report since its June launch, based on entries Google itself has posted to the anomalies page.
| Dates Affected | Report Affected | What Google Said | Resolution |
|---|---|---|---|
| June 24, 2026 | Discover and Generative AI in Discover | A logging error caused a decrease in clicks and impressions on the Discover performance report; reported impressions also dropped for sites with Generative AI features in Discover | No public backfill update found |
| August 13-17, 2026 | Generative AI performance report in Search | A logging error caused a decrease in impressions for that date range | Fixed; Google confirmed missing data was restored on August 21 |
| November 9-11, 2026 | Generative AI performance report in Search | A logging error caused a decrease in impressions; issue affects data logging only | Not yet resolved as of this writing |
Three separate logging errors have hit Google’s Generative AI performance report in its first five months, based on entries Google has published to its own data anomalies page.
Three bugs in five months isn’t a reason to distrust the report outright. New reporting systems built on top of a fast-moving feature like AI Overviews are going to have rough edges. Google has been transparent about each one on a page built for exactly that purpose. My own take is that this is simply what it looks like when a measurement tool gets built at the same speed as the feature it’s measuring. GEO as a discipline doesn’t have mature, stable analytics yet, not from Google and not from the third-party tools trying to fill the gap. That’s worth remembering before you treat any single week of AI impression data as gospel.
The practical habit worth building is small. Before you report an AI visibility drop to a client, a boss, or your own content team, spend two minutes on Google’s anomalies page first. It is a good reason to treat every sudden AI-impression dip as a question first and a crisis second. Build the habit of checking that page into whatever process you already use to review AI visibility each month. Do that consistently, and a week like this one becomes a footnote instead of a scramble.
Frequently Asked Questions
Is the Google AI report logging error fixed yet?
Not as of this writing. Google’s data anomalies page lists the November 9-11 issue without a resolution date. That’s different from the August 2026 bug, where Google posted a confirmed fix and backfilled data about a week after the error began.
Did this bug affect my actual rankings or AI Overview visibility?
No, based on what Google has stated. The anomalies page specifically says the issue affects data logging only. The undercount is in the report, not in how often your pages actually appeared in AI Overviews or AI Mode.
Does the Generative AI performance report cover AI Mode as well as AI Overviews?
Yes. Google’s original June 2026 announcement describes the report as covering citations across both AI Overviews and AI Mode on Search. It groups data by page, country, date, and device, with no breakdown by individual query.
Should I still report a drop to Google through Search Console’s feedback tool?
If your dates match a known entry on the anomalies page, there’s little point, since Google has already confirmed it. If your drop falls outside any listed date range, feedback through Search Console is still the right channel, and it helps Google catch issues the anomalies page hasn’t listed yet.
How often does Google update the anomalies page?
Based on the pattern across all three incidents this year, Google has typically posted a new entry within a day or two of a bug being confirmed. It adds a follow-up note once a fix or data restoration happens, as it did with the August issue.



Leave a response