Google’s May 2026 Core Update and the Preferred Sources Expansion: What Actually Changed

Written by: Khalid SEO | AI Search + SEO Specialist in Malaysia

If your traffic moved around in late May or early June, you weren’t imagining it. Google ran a broad core update from May 21 to June 2, and around the same time, expanded a feature called Preferred Sources into AI Overviews and AI Mode. These are two different things, and mixing them up is the easiest way to spend a week fixing the wrong problem.

I already covered the May 2026 core update as it was rolling out, when things were still moving and there wasn’t a clear picture yet. Now that it’s done and Preferred Sources has landed on top of it, here’s the fuller story: what actually happened, what’s still genuinely uncertain, and what I’d do next if it were my own site.

The May 2026 core update, in plain terms

Google’s core update started rolling out on May 21, 2026, and Google confirmed it was done on June 2. That’s about 12 days, close to the same length as the March 2026 core update, which ran from March 27 to April 8.

What made this one different wasn’t the length. It was the shape of it. Instead of one clean shift, tracking tools picked up several distinct spikes of volatility during the rollout, notably around May 23, again around May 30, and once more in the final 24 hours before Google called it complete. SEO consultant Glenn Gabe, who tracks these updates closely, called the May update considerably more powerful than March, which a lot of practitioners had written off as pretty mild.

A couple of things are worth noting on timing. The update kicked off just two days after Google I/O 2026. A few outlets pointed out how close those two events landed, but Google hasn’t confirmed any real connection between the I/O announcements and the core update itself. Worth flagging, not worth assuming.

Google’s own description of the update was the same boilerplate it always uses: a regular update meant to better surface relevant, satisfying content for searchers. No special guidance, no list of affected sectors. That’s standard for Google, not something specific to May.

One niche took a particularly rough hit. Gambling and other YMYL (Your Money or Your Life) categories saw real volatility, which tracks with how these higher-stakes categories tend to behave during big core updates. If you’re in a YMYL space and you saw big swings, that’s consistent with the broader pattern, not necessarily proof that something’s wrong with your site specifically. I walked through this same logic in more detail after the March update, and the advice hasn’t changed.

If you’re also dealing with the June 2026 spam update that followed close behind, keep the two separate in your head too. Different mechanism, different fix.

When should you actually start analyzing your data?

Not yet, if you haven’t already waited. Google’s own guidance says to wait at least a full week after a core update completes before drawing conclusions, and to compare that week against the week before the rollout started. Since this one wrapped on June 2, the earliest clean comparison window opens around June 9.

The reason for the wait isn’t just Google covering itself. Because this rollout had multiple volatility spikes instead of one smooth curve, a single day’s snapshot from partway through tells you almost nothing. A page that dropped on May 23 might be dealing with something completely different than one that dropped on June 2. Pull your Search Console data too early and you’ll end up chasing noise.

What Preferred Sources actually is (and isn’t)

Separately from the core update, Google announced on May 27, 2026, that it’s bringing a feature called Preferred Sources into AI Overviews and AI Mode.

Preferred Sources isn’t new. It launched in the US and India back in August 2025, after a Search Labs beta earlier that summer. It went global in English in December 2025, then reached all supported languages by the end of April 2026. What changed on May 27 is where it shows up. It used to only affect Top Stories. Now, if a reader has added your site as a preferred source and your content gets cited inside an AI-generated answer, that link carries a visible “Preferred” label, same as it’s worked in Top Stories for months.

Google also rolled out two new carousel formats around the same time. One surfaces timely articles for developing topics. The other surfaces firsthand perspectives pulled from forums, discussions, and social media. And the “Highly Cited” label, which flags original reporting and influential coverage, is showing up in more search results generally now too.

Worth knowing the numbers here. People are reportedly twice as likely to click through to a Preferred Source than to other links, and more than 345,000 unique sources have been selected worldwide so far, up from roughly 90,000 when the feature first went global.

Here’s the part that actually matters for your strategy: Preferred Sources doesn’t touch Google’s ranking algorithm at all. It’s entirely reader-driven. Someone has to go into their settings and add your site themselves. There’s no technical trick, no schema markup, no content restructuring that gets you added as a Preferred Source. Your only real lever is making it easy and obvious for the readers you already have to opt in.

Core update vs. Preferred Sources: what’s actually different

Confusing these two is where a lot of SEO advice is going wrong right now. Rankings and Preferred Sources status answer completely different questions. One is “does Google’s algorithm think this page deserves to rank.” The other is “has this specific reader already decided they trust you enough to follow you around.”

Is GEO or AEO a separate thing you need to learn now?

Short answer: no, at least according to Google, and it’s worth taking that at face value instead of assuming there’s a catch.

Google published new documentation called “Optimizing your website for generative AI features on Google Search,” which builds on guidance it put out in 2025. The line that matters most: from Google Search’s perspective, optimizing for generative AI search is optimizing for the search experience, and is therefore still SEO. Not a parallel discipline. Not something that needs its own toolkit.

I wrote about this exact gap a while back in SEO vs. AEO, and Google’s new guide basically confirms the take I had then. The guide even includes a section literally titled “Mythbusting generative AI search,” where Google lists tactics it says aren’t necessary for its AI features. That list includes llms.txt files, chunking content specifically for AI parsing, and AI-specific schema markup beyond what’s already recommended for standard SEO. Google employees, including Gary Illyes and Cherry Prommawin, have said basically the same thing at Search Central Live events: GEO and AEO don’t need separate frameworks from regular SEO work.

Worth being precise here. This is Google’s stated position for its own AI features, specifically AI Overviews and AI Mode. It doesn’t necessarily carry over to how ChatGPT, Perplexity, or other AI platforms weight things, since those systems aren’t built on Google’s ranking infrastructure and may value different signals. If your AI visibility work also targets those platforms, treat this as one data point, not the whole answer.

For visibility inside Google’s own AI features, though, the takeaway is simple. If your content already ranks well because it’s genuinely useful, well-structured, and shows real expertise, you’re already doing the work that gets you cited in AI Overviews too. You don’t need a separate AI content strategy running next to your regular one.

What to actually do over the next 30 days

1. Hold off on judging the core update’s impact until around June 9. That’s one week after the June 2 completion date, which is the earliest point Google’s own guidance says a comparison will actually mean something.
2. When you do check Search Console, compare the week after completion against the week before the rollout started. Pull specific URLs and queries that moved, not just site-wide averages. A page-level view tells you a lot more than one aggregate number ever will.
3. If you’re in gambling, finance, health, or another YMYL category and you saw big swings, don’t panic and start deleting content. That kind of volatility during major core updates is normal for these categories. Diagnose before you touch anything.
4. Skip llms.txt, skip restructuring your content into AI-specific chunks, and skip speculative AI schema. Google has said plainly that none of this is necessary for its AI features right now.
5. Put a simple prompt somewhere your existing readers will actually see it, encouraging them to add you as a Preferred Source. Google provides instructions and a button template for this. Your newsletter subscribers and returning visitors are the highest-value audience here, since they already trust you.
6. Keep building content that’s genuinely useful and well-sourced. That’s still the thing doing the heavy lifting for both rankings and AI citations, based on what Google itself is saying.

The bottom line

Two things landed close together in early summer 2026: a disruptive core update, and an expansion of a feature that rewards the trust you’ve already built with readers. Neither one calls for a new playbook. The core update rewards the same fundamentals it always has. Preferred Sources rewards the audience relationship you’ve already got. And Google’s own documentation now says, in writing, that optimizing for its AI search features isn’t a separate skill from SEO. It’s the same work with a slightly wider stage.

If you want a second pair of eyes on your Search Console data before you make any changes based on this update, that’s the kind of thing I help with through an SEO audit. And if you’re trying to figure out exactly where you stand in AI Overviews and AI Mode right now, my free tool for tracking AI Overviews gives you a clearer read before you decide what, if anything, to change.

Sources: Google Search Status Dashboard, Google Search Central Blog, Search Engine Land, Search Engine Journal, and Search Engine Roundtable.