---
type: News Edition
title: "SEO, GEO & AI News: May 25th – June 7th, 2026"
description: "Stay ahead with the latest SEO & GEO news (May 25th – June 7th, 2026). Insights on Google's May core update, the llms.txt debate, and Adobe's Semrush deal."
canonical: https://quentinyacoub.com/en/news/seo-geo-ai-news-may-25-to-june-7-2026/
resource: https://quentinyacoub.com/en/news/seo-geo-ai-news-may-25-to-june-7-2026/
lang: en
alternate: https://quentinyacoub.com/fr/actualites/seo-geo-ai-actualites-25-mai-7-juin-2026/
datePublished: 2026-06-08T00:00:00.000Z
dateModified: 2026-06-08T00:00:00.000Z
author: Quentin Yacoub
tags: ["seo news", "geo news", "adobe news", "ai automation news"]
license: CC-BY-4.0
---


> Deterministic, faithful markdown rendering of the canonical page below.
> Not an AI summary. To cite, use the canonical URL.


# SEO, GEO & AI News: May 25th – June 7th, 2026


![News digest cover image](https://quentinyacoub.com/_astro/seo-geo-ai-news-may-25-to-june-7-2026-cover.2Hba_wIi.webp)


This post summarizes key industry shifts from May 25th to June 7th, 2026, focused on SEO and the new era of the agentic web and AI-driven search. Quentin Yacoub breaks down the first measurable hits on llms.txt, ChatGPT query fan-out patterns, Adobe's Semrush acquisition, and Screaming Frog's new MCP server and many more other news, with takeaways for SEO and GEO.


After a few weeks away (parental leave in France, then a week in Basel meeting my Adobe team), I'm back in force with a packed edition of this format that will hopefully satisfy your curiosity about the latest in SEO, GEO, and AI.

The throughline of this edition is the ongoing shift toward the agentic web. We're starting to see the first real hits on llms.txt files, Peec AI breaks down the patterns hiding in ChatGPT's query fan-outs, and Adobe completes its acquisition of Semrush. On the tooling side, Screaming Frog ships an MCP server that changes how you can drive your crawls.

Dive in to see what this means for your SEO and GEO strategies, and jump straight to the sections that impact your daily work the most.

*Have feedback on this digest? [Connect with me on LinkedIn](https://www.linkedin.com/in/quentin-yacoub-seo-geo-specialist/?locale=en_US).*


[Listen to the audio edition](https://quentinyacoub.com/_astro/audio.MQD7RalT.m4a)


## SEO


### What to Do Now That AI Overviews Turned Search Into 4x-Longer Reading Sessions

Source: [Search Engine Land](https://searchengineland.com/ai-overviews-search-reading-sessions-479244) (https://searchengineland.com)

Kevin Indig reports on a clickstream study of ~846,000 U.S. Google sessions showing that when an AI Overview is present, time-on-SERP no longer tracks search intent: all five intent types cluster between 41.9% and 48.5% still on the page at 21 seconds, and sessions run about 4x longer.

**Why it matters:** We already know the average CTR has dropped since AI Overviews landed in Google's results. This study confirms it indirectly, and the takeaway is to adapt so you show up in those results, which at least strengthen brand awareness.

[Read the full study](https://searchengineland.com/ai-overviews-search-reading-sessions-479244)


### Google May 2026 Broad Core Update Is Done Rolling Out

Source: [Search Engine Roundtable](https://www.seroundtable.com/google-may-2026-core-update-done-41435.html) (https://www.seroundtable.com)

Google's May 2026 broad core update has finished rolling out on June 2. Launched May 21, it took under 12 days and Google describes it as a regular, global update meant to better surface relevant, satisfying content from all types of sites, and not a penalty.

**Why it matters:** If you see volatility in your rank-tracking tool, like [Semrush Position Tracking](https://www.semrush.com/position-tracking/), don't panic. Rankings always take a few weeks to settle after a core update, and only then can you really understand what changed.

[Read the documentation](https://www.seroundtable.com/google-may-2026-core-update-done-41435.html)


### Google Drops FAQ Rich Results From Search

Source: [Search Engine Journal](https://www.searchenginejournal.com/google-drops-faq-rich-results-from-search/574429/) (https://www.searchenginejournal.com)

Google has deprecated FAQ rich results, completing a pullback that began in 2023. FAQ structured data can stay on pages without causing issues, but it no longer earns visible FAQ results in Google Search.

**Why it matters:** Adding FAQ structured data has long been a low-effort SEO tactic: it helps Google parse short Q&A content, can earn rich results, and in some cases improves visibility in the People Also Ask feature. Being present for those questions also makes it easier for LLMs to retrieve your pages during their query fan-out. So even with rich results gone, I will keep shipping FAQ markup.

[Read the documentation](https://www.searchenginejournal.com/google-drops-faq-rich-results-from-search/574429/)


## GEO


### How to Make Your Website Agent-Ready

Source: [Suganthan Mohanadasan](https://suganthan.com/blog/how-to-make-website-agent-ready/) (https://suganthan.com)

Suganthan Mohanadasan publishes an evidence-based guide to making a website 'agent-ready' across two surfaces: the page itself (semantic HTML, accessibility) and the protocol layer around it (AI bot rules in robots.txt, llms.txt, the /.well-known/ directory, MCP and A2A cards, WebMCP). His core point is that these protocols don't guarantee AI citations or traffic; what they do is cut agents' fetch-and-parse cost (about 5x smaller payloads) and make your site legible to the systems crawling it.

[Read the full study](https://suganthan.com/blog/how-to-make-website-agent-ready/)


### Patterns in ChatGPT Query Fanouts

Source: [Peec AI](https://peec.ai/blog/patterns-we-see-in-chatgpt-query-fanouts) (https://peec.ai)

Tomek Rudzki analyzes ChatGPT's query fanouts and turns them into optimization guidance. The data shows ChatGPT most often injects words like 'best,' 'top,' 'comparison,' and 'reviews' and pulls review content and current-year freshness even when unasked.

**Why it matters:** A great read. Understanding the query fanouts an LLM runs lets you optimize your pages to rank in the traditional search results those fanouts pull from, so the model retrieves and cites your content instead of a competitor's.

[Read the full study](https://peec.ai/blog/patterns-we-see-in-chatgpt-query-fanouts)


### The Largest llms.txt Audit Yet

Source: [Flavio Longato](https://www.longato.ch/llmstxt-2026-june/) (https://www.longato.ch)

Flavio Longato analyzed 30 days of server logs across thousands of Adobe Experience Manager domains (22,494 requests) to see whether LLMs actually fetch the llms.txt files sites place at their root. Verifiable LLM agents accounted for just 258 hits (1.1% of traffic), while Googlebot was the single most active verifiable crawler with 1,219 requests. The bulk of activity (92.2%) came from unverified tools and AI-readiness auditors rather than language models genuinely consuming the files.

**Why it matters:** Back in August 2025, [Flavio's first study](https://www.longato.ch/llms-recommendation-2025-august/) found essentially zero bot hits on llms.txt files, so seeing measurable traffic now, however modest, tells me the file is starting to register. It also lines up with Google's new Lighthouse agentic browsing score, which checks whether a site ships an llms.txt at all. I read this as early evidence that llms.txt is moving from a debated convention toward something the ecosystem actually monitors.

[Read the full study](https://www.longato.ch/llmstxt-2026-june/)


### Google's Lighthouse Audit Recommends Shipping an llms.txt File

Source: [Chrome for Developers](https://developer.chrome.com/docs/lighthouse/agentic-browsing/llms-txt) (https://developer.chrome.com)

Yet another llms.txt story: Chrome for Developers documents the llms.txt audit inside Lighthouse's agentic browsing category, where Google recommends placing a concise Markdown llms.txt at the domain root so AI agents can grasp a site's structure and key content without crawling the entire architecture. It frames the file as a way to make a site more efficient for agentic exploration.

**Why it matters:** There is an obvious tension here: Google is formally recommending llms.txt through Lighthouse, yet [Flavio's audit](https://www.longato.ch/llmstxt-2026-june/) shows almost no verifiable LLMs actually read the file today. I still think shipping a clean llms.txt is low-effort insurance, since the tooling and recommendations are clearly moving in that direction even if real consumption lags so far.

[Read the documentation](https://developer.chrome.com/docs/lighthouse/agentic-browsing/llms-txt)


### Ahrefs Q1 2026 AI Search Benchmark

Source: [Ahrefs](https://www.linkedin.com/posts/ahrefs_q1-2026-ahrefs-ai-search-benchmark-activity-7465394006324170752-OKAf) (https://ahrefs.com)

Ahrefs published its Q1 2026 AI Search Benchmark, analyzing 146M SERPs and 730K AI responses. The actionable takeaways: YouTube mentions correlate with AI brand visibility more strongly than any other signal, and content length barely matters (53% of cited pages are under 1,000 words). The report also recommends filling information gaps with specific, official content such as FAQs and 'how it actually works' pages, and monitoring brand mentions to correct what AI models say about you.

**Why it matters:** I'm seeing more and more calls to monitor the accuracy of brand mentions inside LLM answers. In classic SEO the holy grail was ranking for non-branded keywords, so that is what we tracked; in the GEO era we also need to track branded queries. [LLM Optimizer](https://business.adobe.com/products/llm-optimizer.html) nailed this by [extracting your brand's mentions across LLMs](https://experienceleague.adobe.com/en/docs/llm-optimizer/using/dashboards/brand-presence) so you can verify the information, [checking YouTube sentiment about your brand](https://experienceleague.adobe.com/en/docs/llm-optimizer/using/dashboards/opportunities/youtube-sentiment-analysis), and [adding relevant FAQs at the bottom of pages](https://experienceleague.adobe.com/en/docs/llm-optimizer/using/dashboards/opportunities/add-relevant-faqs).

[Read the full report](https://www.linkedin.com/posts/ahrefs_q1-2026-ahrefs-ai-search-benchmark-activity-7465394006324170752-OKAf)


### Markdown vs Stripped HTML vs Schema

Source: [Ramp](https://builders.ramp.com/post/marketing-to-ai-agents) (https://ramp.com)

Ramp ran a five-week experiment embedding tracked incentive offers on about 50 marketing pages, served in different formats: markdown, stripped HTML, schema. Markdown was the only format that reliably surfaced in LLM answers, and behavior varied sharply by model.

**Why it matters:** Markdown keeps coming up consistently in the studies I read, even if the measurable wins aren't always huge. Serving your pages as .md is relatively low effort, especially if you run Cloudflare or AEM, so I strongly recommend doing it.

[Read the full study](https://builders.ramp.com/post/marketing-to-ai-agents)


### Beyond RAG: Why Every AI Search Platform Is Now Agentic

Source: [Search Engine Land](https://searchengineland.com/beyond-rag-ai-search-agentic-content-478996) (https://searchengineland.com)

Michael King (iPullRank) explains that AI search no longer works like simple RAG: every major platform now splits your question into many sub-searches (query fan out), picks the right tool for each, and critiques its own draft before answering. For your content, that means going deep across related sub-topics, writing self-contained passages that can stand on their own, and exposing tools or APIs where a function beats an article. He also warns that citation tracking now misses most of your real impact, and shares a practical audit you can run this week.

**Why it matters:** A fascinating article. I especially liked the proposed tracking metrics: sub-query coverage, retrieval-to-citation ratio, reflection survival rate, bridge-entity centrality, tool-call inclusion, and distillation stage-failure rate. I plan to dig into these more and run a few tests.

[Read the full post](https://searchengineland.com/beyond-rag-ai-search-agentic-content-478996)


## Adobe


### Adobe Completes Its Semrush Acquisition

Source: [Adobe](https://news.adobe.com/news/2026/04/adobe-completes-semrush-acquisition) (https://news.adobe.com)

Adobe has completed its acquisition of Semrush, bringing the brand-visibility platform's SEO, GEO, and agentic search optimization (ASO) into Adobe CX Enterprise. Semrush will integrate across Adobe Experience Manager, LLM Optimizer, Commerce, Experience Platform, and Brand Concierge.

**Why it matters:** Data is a critical resource, and Semrush is a leader in that space. This deal should benefit the various AEM tools, and therefore every team that relies on them. As an SEO, I'm genuinely excited about this acquisition. Welcome to the Semrush teams.

[Read the announcement](https://news.adobe.com/news/2026/04/adobe-completes-semrush-acquisition)


### Adobe: The Decisions That Determine Whether AI Scales Value or Risk

Source: [Adobe](https://business.adobe.com/blog/decisions-that-determine-whether-ai-scales-value-or-risk) (https://business.adobe.com)

Emily McReynolds, Adobe's Head of Global AI Strategy, offers a framework to vet AI partners through four lenses: data trust, output control, governance over agent actions, and brand-safe quality at scale. She notes 65% of organizations had an AI agent-related security incident in the past year, and gives sample due-diligence questions for each lens.

**Why it matters:** I fully agree with Emily: we keep handing AI more responsibility and more access, both at work and in our personal lives. I see AI governance as an essential issue to build into our strategies, and I'm glad Adobe is positioning itself at the forefront here.

[Read the full post](https://business.adobe.com/blog/decisions-that-determine-whether-ai-scales-value-or-risk)


## Tools


### Screaming Frog Adds an MCP Server

Source: [Screaming Frog](https://www.screamingfrog.co.uk/blog/seo-spider-24/) (https://www.screamingfrog.co.uk)

Screaming Frog's latest update introduces an MCP server that lets you drive the crawler with natural language from AI assistants. Through the MCP you can run and summarize crawls, manipulate and visualize the data, and automate export, data-combining, and analysis pipelines.

**Why it matters:** I've been waiting for this for a long time. You could already ask Claude, Codex, or another tool to analyze a page or a few, but it was never fully reliable. Screaming Frog is my favorite SEO tool, and being able to trigger a crawl from Claude Code and move straight into the analysis is fantastic; if your site is also connected to Claude Code, you can even fix the issues right away. Honestly, it's wild. Thank you, Screaming Frog.

[Read the documentation](https://www.screamingfrog.co.uk/blog/seo-spider-24/)


### Search Console Adds Generative AI Performance Reports

Source: [Google Search Central](https://developers.google.com/search/blog/2026/06/gen-ai-performance-reports) (https://developers.google.com/search)

Google launched new generative AI performance reports in Search Console that show how a site appears in AI Overviews and AI Mode. The reports surface impressions, pages, countries, devices, and dates, and the same data also stays in the overall Performance report.

**Why it matters:** This gives us a clearer picture of the prompts we rank for, plus real data on the keywords and prompts that trigger AI Overviews or get used in AI Mode. Unfortunately it doesn't provide clicks or the answer it gives for the prompts.

[Read the announcement](https://developers.google.com/search/blog/2026/06/gen-ai-performance-reports)


### Chrome's Lighthouse Adds an Agentic Browsing Score

Source: [Chrome for Developers](https://developer.chrome.com/docs/lighthouse/agentic-browsing/scoring) (https://developer.chrome.com)

Chrome for Developers documents a new experimental Lighthouse category that scores how well a site is built for AI agent interaction. It runs deterministic audits across several areas: WebMCP integration to expose tools to AI agents, accessibility-tree health (the machine's primary data model), layout stability for reliable element interaction, and the presence of an llms.txt file at the domain root.

**Why it matters:** I see this as another step in Google's transition toward the agentic web, and the llms.txt check is its most contested part. A few months ago, server logs showed virtually no bot hits on the file; [Flavio Longato's recent study](https://www.longato.ch/llmstxt-2026-june/) of 5,000 sites now finds some, mostly from Google. Lighthouse formally scoring llms.txt gives weight to a standard the SEO community is still debating.

[Read the documentation](https://developer.chrome.com/docs/lighthouse/agentic-browsing/scoring)


### 7 Ways to Automate Content Marketing with Ahrefs' Agent A

Source: [Ahrefs](https://ahrefs.com/blog/agent-a-for-content-marketing/) (https://ahrefs.com)

Ryan Law, Ahrefs' Director of Content Marketing, walks through Agent A, an autonomous marketing agent with unrestricted access to the full Ahrefs dataset. He details seven content-marketing workflows it powers: an 11-stage blog-writing pipeline, an article-updating pipeline with side-by-side diffs, automated monthly performance reports, a blog topical-authority audit, a competitor-content feed, a LinkedIn swipe file, and scientific internal-linking recommendations.

[Read the full post](https://ahrefs.com/blog/agent-a-for-content-marketing/)


---
**Cite:** https://quentinyacoub.com/en/news/seo-geo-ai-news-may-25-to-june-7-2026/
**License:** CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/), Quentin Yacoub
**Related:** [SEO, GEO & AI News: June 8th – June 21st, 2026](/en/news/seo-geo-ai-news-june-8-to-june-21-2026.md) · [SEO, GEO & AI News: April 13th – April 26th, 2026](/en/news/seo-geo-ai-news-april-13-to-april-26-2026.md) · [SEO, GEO & AI News: April 6th – April 12th, 2026](/en/news/seo-geo-ai-news-april-6-to-april-12-2026.md) · [SEO, GEO & AI News: March 23rd – April 5th, 2026](/en/news/seo-geo-ai-news-march-23-to-april-5-2026.md)
**See also:** /llms.txt · /okf/index.md