This article was AI-generated as part of an experimental historical-content project. The date reflects the period being analyzed rather than the date the article was originally written.
Google released the June 2025 core update on Monday and said the rollout may take up to three weeks. It came with no new advice. As Search Engine Land noted, it is the second core update of the year, arriving just over three months after March’s.
The routine parts are familiar. Rankings will move, tracking tools will show volatility, and readings taken mid-rollout are unreliable. I want to look instead at a detail of timing that changes what a core update means for anyone watching how a company is described.
The first core update with AI Mode open to all
When the March update rolled out, AI Mode was an opt-in experiment in Labs. On May 20 Google began rolling it out to everyone in the U.S. as its own tab in Search. This is the first core update since.
That matters because of how Google describes the link between its AI features and its rankings. When it introduced AI Mode in March, Google wrote that it is “rooted in our core quality and ranking systems,” with additional techniques drawing on the model’s reasoning to improve factuality. In May, John Mueller of Google told site owners that “the underpinnings of what Google has long advised carries across” to AI Overviews and AI Mode, and that meeting Search’s technical requirements covers AI formats as well.
Put those statements together and the implication is plain. A core update changes the quality and ranking systems. AI Overviews and AI Mode draw on those systems to decide which pages to retrieve and cite. So a core update is also, indirectly, a change to the material generative answers are built from.
A case worth reasoning through
Take a company with an ordinary profile: its own site, a Wikipedia article, a handful of news stories including a critical investigation from two years ago, some review sites and a few forum threads. Before Monday, a particular mix of those pages ranked for the company’s name and for common questions about it.
Suppose the update reassesses forum content, or older news, for those queries. Two things can now shift together. The order of blue links changes, which reputation teams have always tracked. And the pool of pages that AI Overviews and AI Mode consider for the same questions changes too. The sentences in a generated answer, and the sources cited beside them, may change without anything new having been published about the company.
That second effect is easy to miss. Most teams track rankings by URL. Few have a record of what the AI Overview said about them, or which sources AI Mode cited, before Monday. Without that baseline, a changed answer next month will look like a new problem rather than a downstream effect of a ranking change.
I don’t want to overstate this. Google has not said that core updates change AI answers in any specific way, the generative layer has its own models, and answers vary from one query to the next regardless. But the coupling is Google’s own description, and it is a reason to treat this update differently from earlier ones. Core updates have long rewarded substance over performance for the algorithm. They now decide, in part, what the answer box reads.
What to do while it rolls out
Capture now. For the names and questions that matter most, the company, its executives, its main products and any controversies, save the classic results page, the AI Overview and the AI Mode response with its citations, and date each one. Repeat after Google says the rollout is complete, and again a week later.
Then compare by source type, not just by URL. Did forum threads gain ground in both layers? Did an old investigation drop from the links but stay in the answer, or the reverse? Those patterns are the actual case study. The update is just the occasion for running it.