Google added AI search visibility reporting to Search Console. Not a new product. Not a "Generative Console." The same dashboard you already use to check rankings now shows how often your pages appear in AI Overviews, AI Mode, and Discover's AI features. The rollout started with UK sites in early July 2026 and is expanding.
Where Google put the button is the argument. If AI visibility required a separate discipline, it would have gotten a separate tool. It didn't.
What the report actually shows (and what it doesn't)
The new generative AI performance reports track impressions: how often your pages surfaced inside Google's AI features. You get the same dimensions as the standard performance report. Pages, countries, devices, dates, down to hourly granularity.
Two absences matter more than anything included. First, there are no clicks. You can see that you appeared in an AI Overview. You can't see whether anyone acted on it. Presence, not consequence. Second, the release ships alongside a control that lets you opt your content out of AI responses entirely. Google handed operators a meter and an exit switch in the same update. That pairing is deliberate.
For demand gen teams, this creates a measurement problem worth naming: impressions are a leading indicator, not a pipeline signal. A free impressions report is the most seductive version of the "AI visibility equals business result" trap, because the cost of pulling it drops to zero. Easy to chart, easy to drop in a deck, disconnected from whether a buyer converted.
The streetlight effect is about to distort your reporting
The moment AI visibility becomes free and native inside a tool every operator already has open, it becomes the AI visibility people actually watch. Not because Google's surface matters more than ChatGPT, Claude, or Perplexity. Because it's the one with a free dashboard.
This is the streetlight effect applied to an entire channel. You look for your keys under the lamppost because that's where the light is. Google just switched on a bright, free light over its own surface. The darker corners, where standalone trackers charge you to look, get less attention by default.
The trouble: AI visibility is plural. Most AI-cited pages appear in only one engine. A page cited constantly in one model can be absent from the next. A Google-only view is one engine out of several, handed to you with the authority of a number in a tool you already trust. The research backs this up: the median brand is cited in only 3% of AI Overviews even when ranking organically. AI Overviews appear in roughly 50% of relevant B2B searches. That gap between organic rank and AI citation is real, and it exists across engines, not just Google's.
Cross-engine trackers aren't the losers here. They do the harder thing Search Console never will. The risk sits on the operator's side: a free, native, single-engine number is easy to over-trust.
What this means for your SEO reporting cadence
If you've been buying a separate GEO practice with its own budget line and quarterly report, Google just disagreed in the most concrete way a platform can. AI visibility is search visibility. Fold it into the SEO reporting cadence you already run.
The practical moves:
- Pull the generative report for the Google slice when it reaches your account. Treat AI Overview impressions the way you treat standard impressions: a sign you were eligible, not proof you won.
- Keep one cross-engine check for the surfaces Google will never show you. ChatGPT, Claude, Perplexity each have their own citation logic. Someone still has to see those.
- Track citation inclusion rate as a KPI alongside traditional rank visibility. This is emerging as the metric that separates "we rank" from "we get cited."
- Make the opt-out a deliberate decision. Don't discover it later as a default you missed.
The risk of ignoring the multi-engine picture is real. Reports suggest 20-50% of traditional organic traffic may be at risk from AI interception. Enterprise leads from AI sources have been reported as 43% more qualified than traditional channels. That's a signal worth measuring across every surface, not just the free one.
The feature is useful. The placement is the point.
Google filed AI visibility under search because, as far as Google is concerned, that's where it has always lived. Search Console has defined what counts as search performance for twenty years. The thing it reports is, by definition, the thing Google treats as search performance. AI Overviews showing up next to blue-link impressions is an accounting decision, not a product launch.
The takeaway isn't a tool to learn. It's a budget line to kill. And then the harder work begins: building the measurement layer that covers the engines Google will never report on, and connecting all of it to pipeline instead of impressions. The free dashboard is the starting line. Mistaking it for the finish is the trap.
