SEO News Roundup: The Week Google I/O Changed the Conversation (Again)

Week of May 19–25, 2026

Every year around Google I/O, my LinkedIn feed turns into a disaster movie. “SEO is dead.” “The ten blue links are gone.” “Your traffic is about to collapse.” I’ve been in this industry for 16 years, and I’ve seen this same cycle play out so many times I could write the headlines before the keynote ends.

This week was no different — except that some of the concern is actually warranted. Just not for the reasons most people are yelling about.

Here’s what actually happened this week, and what I think it means.

1. Google I/O Didn’t Kill SEO — But It Did Change the Economics

The biggest story this week wasn’t a ranking update. It was Google I/O 2026 and the avalanche of takes that followed.

Google announced a redesigned Search box that accepts images, files, videos, and Chrome tabs alongside text. Gemini 3.5 Flash became the global default AI model. AI Mode quietly crossed one billion monthly users, with queries doubling quarter over quarter. And the headline feature: “information agents” that monitor the web on your behalf and alert you when something relevant changes — an apartment listing, a price drop, a news story.

The hot takes were predictable. TechCrunch declared “Google Search as you know it is over.” LinkedIn posts echoed an “SEO is dead” refrain within hours of the keynote. Google pushed back directly on X the next day, clarifying that AI Mode is not the default experience and that users still get a range of results.

Here’s my honest read after 16 years of watching these cycles: both sides are wrong.

The “SEO is dead” crowd is overstating what Google announced. Blue links aren’t gone. The web tab still exists. Google’s own optimization guide literally calls AEO and GEO “still SEO”. These systems still depend on crawling, indexing, and ranking the web — the same foundations they’ve always used.

But the “calm down, nothing has changed” crowd is underselling the real issue. The risk from I/O isn’t that links disappear. It’s that fewer people need to click them. AI Overviews have already cut organic clicks on triggered queries by 38% in field experiments. The average AI Mode query is now three times longer than a traditional search. Planning queries grew 80% faster than other types. People are delegating more of their research process to Google — and information agents are the next step in that direction.

A tool that monitors the web for you and delivers synthesized updates inside Google? That’s not a visit to your site. Your content gets consumed; you just don’t see it in your analytics.

The risk is economic, not technical. And that’s a harder conversation to have.

What I’d watch: If you’re tracking organic performance, now is the time to segment your content by query type. Simple-answer content — store hours, return policies, basic “what is” explainers — is the most exposed category right now. Original research, proprietary data, and expertise that AI can’t synthesize elsewhere? That’s where your defense lives. If you haven’t done a content audit recently, our free SEO audit template is a good starting point for categorizing your content by risk.

Sources: Search Engine Journal | Search Engine Land

2. May 2026 Core Update Is Now Rolling Out

While everyone was processing I/O fallout, Google quietly dropped the May 2026 core update on May 21 — the second broad core update of the year.

No companion blog post. No specific targets announced. Just a dashboard entry that says “released the May 2026 core update, rollout may take up to 2 weeks.” Classic.

The March 2026 core update wrapped up on April 8 after 12 days. We covered what that fallout looked like in April — the pattern of sites seeing dramatic ranking shifts that took weeks to stabilize. The same caution applies now.

This one lands at a genuinely chaotic moment. Sites are already processing I/O announcements, AI Mode behavior changes, and the general uncertainty of what “good content” means in an AI-first world. Throwing a core update into that mix is going to make it really hard for people to diagnose what’s affecting their traffic.

My advice: Don’t panic-publish or start rewriting pages based on early signals. Wait until the rollout completes and give it another week before pulling conclusions from Search Console. Look at pre-May 21 performance as your baseline. Core updates are not targeted at specific content types — pages move up and down as Google recalibrates across the web.

If you saw drops during the March update that never recovered, this one is worth watching closely. Back-to-back updates within the same quarter can compound. Not sure how to diagnose what’s happening to your site? Our small business SEO audit guide walks through exactly that process.

Sources: Search Engine Journal | Search Engine Land

3. Google Dropped FAQ Rich Results (Official Confirmation)

This one flew under the radar amid all the I/O noise, but it’s worth flagging again as the timeline becomes official.

As we covered in last week’s roundup, Google officially killed FAQ rich results in search as of May 7th. The full deprecation timeline:

• May 7: FAQ rich results stop appearing in search entirely
• June: Google removes the FAQ filter, rich result report, and Rich Results Test support from Search Console
• August: API support for FAQ search appearance gets axed

If you built strategy around FAQ schema for the expanded accordion snippets in the SERP — that real estate is gone. The markup itself won’t hurt you if you leave it, but stop building around it.

Source: Search Engine Journal

4. WordPress 7.0 Lands With Native AI Integration

WordPress 7.0 shipped this week with native AI baked into the core platform — and the reaction has been mixed, but I think it’s a net positive for most site owners.

This is being called the most consequential CMS release in years. The AI integration goes beyond plugin territory. For content teams and small business owners, that changes what’s possible without needing to stitch together third-party tools.

There’s a security flag worth noting: a researcher found that WordPress 7.0’s AI features could create a new attack surface for stealing AI API keys if sites aren’t configured carefully. If you’re running WordPress at scale, make sure your hosting setup and plugin stack are updated to handle this before enabling the new AI features.

For individual site owners? The upside is real. Just vet your setup first.

Source: Search Engine Journal

5. Microsoft Clarity Now Shows Grounding Queries Behind AI Citations

This is the quiet tool update that deserves more attention than it’s getting.

Microsoft Clarity now surfaces the grounding queries behind AI citations — meaning you can start to see how AI engines are decomposing user intent when they pull from your content. Their research suggests this logic is platform-agnostic, so you’re getting a window into how AI systems generally think about queries, not just Bing’s.

For anyone trying to figure out how to optimize for AI citation eligibility — what some are calling Answer Engine Optimization (AEO) — this kind of visibility is genuinely useful. Instead of guessing what queries your content is being pulled for in AI answers, you have data.

I’m not ready to build an entire strategy around this yet. The sample sizes are still small and the behavior is evolving fast. But I’d absolutely set it up on any site I’m actively working on and start collecting the data now, even if you don’t know what to do with it yet. Data you collect today will matter a lot more when the signal is clearer in 6–12 months.

Source: Search Engine Journal

6. What Makes a Brand “Machine-Readable” in AI Search?

Search Engine Land ran a piece this week that I thought was genuinely useful amid all the noise: what makes a brand machine-readable in AI search?

The argument is that structured brand signals — clear entity relationships, consistent mentions across authoritative sources, well-defined product and service descriptions — are increasingly what separates brands that get cited in AI answers from those that don’t.

This connects to something I’ve been saying for a while: brand building and SEO have never been more aligned. If an AI model doesn’t have a clear picture of what your brand does, who it serves, and why it’s credible, you’re invisible to it. That’s not a technical problem — it’s a content and PR problem.

The practical implication for small businesses: Make sure your Google Business Profile is complete, your schema markup clearly defines your entity, and your brand is consistently mentioned in the same way across your website, directory listings, and any press coverage. This is exactly the kind of work our DIY SEO guide covers for businesses who want to build this foundation without an agency.

Source: Search Engine Land

7. LLM Optimization Guidance Is Still a Mess — And That’s Okay

Former Bing search lead Duane Forrester wrote something worth reading this week: LLM optimization guidance doesn’t transfer the way SEO guidance used to.

His point is that with traditional search engines, there was shared understanding — what worked on Google largely worked on Bing. The industry could develop common frameworks. With LLMs, that portability doesn’t exist. Each model learns differently, updates differently, and hasn’t bought into any shared standard for what “optimization” means.

I think this is right, and it’s one of the more honest assessments I’ve seen from someone with actual insider experience. Anyone selling you a “guaranteed LLM optimization strategy” right now is guessing. Including me.

What I’d recommend instead: focus on fundamentals that have always worked — authoritative sourcing, clear entity signals, original expertise, structured content that’s easy to parse. Those translate across systems even when specific tactics don’t.

If you’re spending time on internal linking and content architecture right now (a great use of your time), our Internal Link Recommender tool can surface the opportunities you’re missing without spending hours doing it manually.

Source: Search Engine Journal

The Week in Summary

This was a week that felt heavier than most. Between the Google I/O aftermath, a core update mid-cycle, FAQ rich results officially gone, and WordPress reshaping what a CMS looks like in the AI era — there’s a lot to process.

My honest take after 16 years: the fundamentals still matter. Create content that’s genuinely useful, build brand authority that AI systems can recognize and cite, and don’t make reactive moves based on 48-hour hot takes.

The SEO industry has survived every “SEO is dead” moment by adapting rather than abandoning what works. This one is no different — just noisier.

If you want a hand making sense of how your specific site is positioned heading into this environment, book a consultation and let’s talk through it.

See you next Monday.

Your Friendly SEO is a weekly roundup written from the perspective of someone who has been doing this work for 16 years. Subscribe to the newsletter so you don’t miss next week’s issue.