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 May 2026 core update this morning, U.S. Pacific time. The dashboard entry is the usual single line: “The rollout may take up to 2 weeks to complete.” In a LinkedIn post quoted by Search Engine Land, Google called it “a regular update designed to better surface relevant, satisfying content for searchers from all types of sites.” It is the second broad core update of the year, after March.
It also lands two days after I/O, where Google made Gemini 3.5 Flash the default model in AI Mode and redesigned the Search box. Most people will file these as separate stories, one for SEO teams and one for everyone else. Mechanically they are separate. For anyone monitoring how a company appears in Search, they meet in the same place.
Two different kinds of change
A core update changes ranking systems: the machinery that decides which pages in Google’s index are most relevant and useful for a query, and in what order. A model change, like the one announced on Tuesday, changes the system that reads those pages and writes the answer. One decides what gets considered. The other decides how it is phrased.
Generative features sit on top of both. Google has been explicit about this since the early days of AI Overviews. In a May 2024 post about errors in the feature, Elizabeth Reid wrote that the model behind AI Overviews “is integrated with our core web ranking systems,” designed to identify “relevant, high-quality results from our index,” and built “to only show information that is backed up by top web results.” She attributed most mistakes not to invention but to misread queries, misread nuance, or “not having a lot of great information available.”
Google’s newer guidance points the same way for AI Mode. Its guide on generative AI features, published last week and summarized by Search Engine Journal, says a page must be indexed and eligible to show a snippet in Search to appear in AI Overviews or AI Mode.
So if an AI answer is grounded in top results, anything that reorders top results can change the answer, even when the model is identical. A core update is also an update to the evidence AI answers draw on.
Why this matters for a company name
For most brand queries, the page that matters is not the one ranked first. It is the third or fifth result: the critical article, the outdated review, the forum thread that ranks well enough to be retrieved. A modest movement in those positions, the kind core updates routinely cause, can decide whether that page is among the sources an AI answer is built from. The generated paragraph can change in tone without a single new article being published.
Google’s advice to site owners during core updates says nothing about generative features. That is reasonable from Google’s side. For reputation teams it means the translation has to happen internally.
What to record over the next two weeks
For a set of company and executive queries, capture both the classic results and the sources cited in any AI Overview or AI Mode answer. Keep the two lists side by side and dated. Note the date the model default changed, May 19, and the date the core update began, May 21, so later changes can at least be placed in time.
Do not draw conclusions mid-rollout. The point of collecting now is to have a record when the rollout ends, rather than reconstructing one from memory.
Earlier in this archive I wrote that “helpful content” isn’t a reputation strategy. The same logic applies here. Ranking systems reward what they judge to be useful sources, and AI answers inherit that judgment. In a core update week, a company’s best protection is still to be the most useful source about itself.