Welcome to the week’s Pulse: This week’s updates touch rankings, the Search interface, AI Mode behavior, and Google’s guidance around AI agents.
Google launched a core update, announced what it called the biggest upgrade to the Search box in over 25 years, released first-party AI Mode usage data, and sent mixed signals on llms.txt from two different product teams.
Here’s what matters for you and your work.
Google Begins Rolling Out The May 2026 Core Update
Google began rolling out the May core update on the 21st, per a message on the Google Search Status Dashboard.
Key facts: This is the second Search core update of 2026 and the fourth confirmed ranking update this year. The rollout may take up to two weeks. Google hasn’t published a companion blog post or shared goals for the update.
Why This Matters
The timing puts this update in the middle of Google I/O week. Ranking movement over the next two weeks will overlap with other changes Google announced, which could make it harder to isolate what caused any shifts you see in Search Console.
Your baseline should be the weeks before May 21, compared against performance after the rollout finishes. Wait at least one full week after completion before reviewing data.
What SEO Professionals Are Saying
Marie Haynes, founder of Marie Haynes Consulting Inc., connected the timing to I/O:
“Makes sense seeing as Gemini 3.5 Flash is now powering the AI features of Search.”
Harpreet Singh Chatha, SEO & AI Search Consultant, suggested this update could be targeting websites that are over-optimizing for AI citations:
“Calling it now. If you’ve been doing dumb [things] to show up in AI answers this one’s coming for you.”
Read our full coverage: Google Begins Rolling Out May 2026 Core Update
Google Redesigns Search Box, Upgrades AI Mode & Previews Search Agents At I/O
Google announced Gemini 3.5 Flash as the new default model in AI Mode, redesigned the Search box with AI capabilities, and previewed information agents coming this summer.
Key facts: Google described the redesigned Search box as its biggest upgrade in over 25 years. It expands dynamically, supports multimodal inputs like images and files, and provides AI-powered suggestions beyond autocomplete. Information agents will monitor the web and deliver updates. New features include agentic booking, generative UI, and Personal Intelligence expansion to nearly 200 countries.
Why This Matters
The Search box redesign prompts users to describe needs in longer, conversational queries. Paired with Gemini 3.5 Flash in AI Mode, responses mostly come from AI rather than traditional pages.
Information agents continuously search and synthesize updates, raising questions about whether your content is cited or overlooked in those summaries.
What SEO Professionals Are Saying
Jake Ward, SEO/content entrepreneur building Mentions, wrote:
“My feed is full of ‘SEO is dead’ posts today. But none of this should be a surprise. Every platform we use is moving towards AI-first, agentic, proactive experiences like this. And clicks were already dying. We’ve watched CTR decline for 2-3 years straight now. However, search is very much alive, just different. We’re moving further into a world of visibility > clicks.”
Read our full coverage: Google’s New Search Box Hands Queries To AI Agents, I/O Reveals
Google Releases First AI Mode Usage Data After One Year
Google published a report on how people use AI Mode in the U.S., drawing on internal Search data and Google Trends one year after launch.
Key facts: AI Mode has over 1 billion monthly users, with queries doubling each quarter. Searches are thrice as long as traditional ones, and follow-up queries increase 40% monthly in the U.S. Over 16% of searches are multimodal, using voice, images, or video. Planning queries grow at 80% the rate of overall usage. Trends data isn’t publicly available.
Why This Matters
The key insight is the behavioral data, not milestone numbers. Users write longer queries, follow up more, and use multiple input types, changing content and how it surfaces.
Pages with short keywords may not match AI Mode’s conversational patterns. Growth in planning queries is significant; when users ask AI Mode to compare products, evaluate services, or research, that content has commercial value, even without clicks.
Google’s report is based on internal, unverifiable data, and AI Mode search trends aren’t available publicly.
What SEO Professionals Are Saying
Jeffrey Cohen, chief business development officer at Skai, wrote:
“Shoppers aren’t typing ‘running shoes.’ They’re asking ‘what are the best running shoes for a wide foot that I can wear for a half marathon training on pavement under $150.’ That’s not a keyword. That’s a brief. Planning queries grew 80% faster than AI Mode overall in the last 6 months. That means shoppers are using AI as a research partner long before they hit buy. The brand that shows up during research owns the consideration phase. The transition from keyword to conversation has been talked about for years. The data says it’s already here.”
Alisa Scharf, CAIO at Seer, pointed to the measurement gap:
“For those of us who weren’t yet investing in tracking visibility in AI Mode, we’ve gotta bug our product or procurement teams to get more tracking established. Can’t optimize what you can’t measure. If anyone from Google follows me — It’s really pretty wild that none of these metrics are available for free in Google Webmaster tools. I so rarely shake my fist in the general direction of Palo Alto, but this is becoming obscene. Big props to the team at Bing who’s investing in a real control center of information with their Webmaster Tools.”
Read our full coverage: Google Reveals First AI Mode Usage Numbers After One Year
Google’s Llms.txt Guidance Splits Between Search & Lighthouse
Google’s Search team and its Lighthouse team are giving different guidance on llms.txt. Meanwhile, Mueller clarified where markdown pages for LLMs do and don’t help.
Key facts: Google’s AI guide states llms.txt isn’t needed for AI Search. Lighthouse 13.3 checks for llms.txt by default, flagging sites with errors. Mueller said markdown pages are useful for documentation but not for most websites. He differentiated discovery (search visibility) from on-page tasks, advising sites to focus on being found.
Why This Matters
The answer depends on your traffic. If agentic tools visit and complete tasks, markdown versions of your docs may help those tools. If you’re planning for the future, Mueller recommends prioritizing current needs first, saying, “Prioritize needs before dreams.”
The conflict between the Search and Lighthouse teams exists, and Google hasn’t resolved it. Search Central offers guidance for Search visibility, while Lighthouse assesses agentic browsing readiness, not ranking or AI Mode eligibility.
What SEO Professionals Are Saying
Chris Long, co-founder of Nectiv, wrote:
“Chrome just released documentation on their new Agentic Browsing audits. Somewhat buried in there, they reference how the audit will check for the LLMs.txt file. It mentions how it ‘checks for the presence of a machine-readable summary at the domain root.’ This is less than a week after their documentation talking about how SEOs don’t need to worry about additional files + markup. I’m starting to turn around on the LLMs.txt a bit. Some seriously smart people including Crystal Carter, John-Henry Scherck, Joost de Valk are turning me around on it a bit. It also seems very clear that Google doesn’t want us to test this stuff. To just keep doing SEO as we’ve normally done it without looking behind the scenes and waiting for their guidance.”
Read our full coverage: Mueller Explains Why Google Uses Markdown On Dev Docs | Google’s llms.txt Guidance Depends On Which Product You Ask
Theme Of The Week: The Quiet Rebuild
Google is rebuilding Search around AI while telling everyone the fundamentals still apply.
The optimization guide published last week says AEO and GEO are “still SEO.” Mueller says to focus on current needs. The core update rollout looks like any other. But the same week, Google announced what it called the biggest upgrade to the Search box in over 25 years, reported that AI Mode has surpassed 1 billion monthly users, announced always-on information agents, and released data showing user behavior is already shifting toward longer, multimodal, follow-up-heavy queries.
The gap between Google’s public guidance and its product roadmap keeps widening. The infrastructure is changing faster than the guidance.
Top Stories Of The Week:
- Google Begins Rolling Out May 2026 Core Update
- Google’s New Search Box Hands Queries To AI Agents, I/O Reveals
- Google Reveals First AI Mode Usage Numbers After One Year
- Mueller Explains Why Google Uses Markdown On Dev Docs
- Google’s llms.txt Guidance Depends On Which Product You Ask
More Resources:
- Google’s New AI Search Guide Calls AEO And GEO ‘Still SEO’
- Google Expands SynthID Verification To Search
- Google’s Mueller On Why AI-Built Websites Miss SEO Basics
Featured Image: PeopleImages/Shutterstock
