Google Just Put an AI Agent Between Your Brand and Every Searcher. One Billion of Them.

On May 19, Google announced the biggest upgrade to Search in 25 years. AI Mode is now the default for one billion monthly users. Persistent information agents, rolling out this summer, will monitor topics continuously and select sources on the user's behalf without requiring a second query. If your brand is not in the source set those agents pull from, you are not in the answer. And in Ahrefs' analysis of Google AI Overviews, only about 38% of cited pages also ranked in Google's top 10 for the same query.

That last number should stop you cold. Three out of four brands sitting in traditional organic positions one through ten are completely invisible to the system that now stands between them and the person searching.

The Search Box Is No Longer a Search Box

Google rebuilt the search interface from scratch. The new box accepts images, PDFs, files, videos, and full conversational prompts instead of keyword fragments. It anticipates intent before typing finishes. Gemini 3.5 Flash, Google's newest model, now powers AI Mode globally, handling multi-step reasoning and maintaining context across longer interactions.

This is not an incremental feature release. The interface itself has been redesigned around the assumption that users will ask questions, not type keywords. AI Overviews now reach 2.5 billion monthly active users across 200+ countries. The answer layer is the primary surface.

The old model was rank, click, session, convert. The new model is visibility, influence, outcome. Seer Interactive found that organic click-through rates on informational queries with AI Overviews fell from 1.76% in June 2024 to 0.61% in September 2025 (3,119 queries across 42 client organizations). SparkToro's 2024 study found that 58.5% of US Google searches ended with no click.

The click is not dying. It is already dead for most queries. The question is whether your brand makes it into the answer before the click even becomes an option.

Information Agents Change the Game Again

Here is the part most people will miss. Google's new persistent information agents do not wait for a query. They run continuously in the background, tracking topics the user cares about: apartment listings, competitor pricing, product launches, industry shifts. The agent decides which sources to pull from. It decides what matters. It delivers updates without the user lifting a finger.

Think about what that means for your brand. The user is no longer the one searching. The agent is. And the agent is not browsing your website, scanning your blog, or reading your LinkedIn posts. It is pulling from a source graph built on structured data, entity signals, citation history, and freshness.

If your content is not structured for machine extraction, the agent will never select it. Not because it failed some ranking algorithm. Because it was never legible to the system in the first place.

The Source Architecture That Agents Actually Use

The data on what gets selected and what gets ignored is precise.

Only 30% of brands remain visible from one AI answer to the next. Just 20% stay present across five consecutive runs. This is not a traffic fluctuation. It is a structural visibility lottery where most brands lose every round.

What determines whether you stay in or fall out:

Freshness is non-negotiable. Pages not updated quarterly are 3x more likely to lose citations.

Third-party pages carry most brand discovery. 85% of brand mentions originate from third-party pages, not owned domains. In AirOps' October 2025 offsite-signals report — one vendor's sample of top-of-funnel commercial queries — brands were 6.5x more likely to be mentioned through third-party sources than their own sites. 48% of AI citations come from community platforms like Reddit and YouTube.

Structure determines extractability. Sequential headings correlate with 2.8x higher citation likelihood. 61% of cited pages use three or more schema types.

Core Web Vitals are hygiene worth checking. Slow or unstable pages are harder for crawlers and visitors alike, and no study we could trace to its publisher measures a citation penalty for failing them.

This is the source architecture that information agents will use to decide whether your brand exists in their answers or not.

Why This Is a Machine Relations Problem, Not an SEO Problem

I have spent eight years placing brands in front of the audiences that matter most. I built AuthorityTech around the thesis that the relationship between brands and machines would become the most important relationship in business. Google just proved it.

Traditional SEO was built for a world where humans browsed pages. Machine Relations is built for a world where machines select sources. The distinction is not semantic. It is structural.

When 60% of Americans are reading AI summaries instead of clicking through to websites, and when AI Overviews recommend your competitors 69% of the time, the question is not whether you rank. The question is whether the machine considers you a credible source worth citing.

And the economics are clear. Semrush's study put the average AI search visitor at 4.4 times the value of the average organic visitor, by conversion rate. The brands that build for machine source selection are not just visible. They are reaching visitors who converted above organic in the studies we cite.

The brands that do not are watching from outside the conversation while information agents recommend someone else.

What to Do Right Now

Three moves. All of them are concrete.

1. Audit your AI visibility today. Go to ChatGPT, Perplexity, and Google AI Mode. Search the queries your buyers actually use. Do not search your brand name. Search the problem you solve. If you are not in the answer, your page-one ranking is a vanity metric.

2. Build for extraction, not just ranking. Answer the query in your first two sentences. Use sequential headings. Add schema markup. Keep your Core Web Vitals clean. Refresh quarterly at minimum. These are not nice-to-haves. They are the mechanical gates that determine whether an AI agent can even read your content.

3. Earn third-party citations as if your visibility depends on it. It does. When 85% of brand mentions come from third-party pages, your owned content strategy alone will not get you into AI answers. Earned media, community presence, and independent validation are the signals that machines trust. That is the core of Machine Relations.

FAQ

What are Google's information agents in AI Mode?

Information agents are persistent AI systems announced at Google I/O 2026 that run continuously in the background, monitoring topics and selecting sources on behalf of the user. Unlike traditional search, they do not require a query to activate. They proactively deliver updates from sources they deem credible and relevant.

How many people use Google AI Mode?

As of the I/O 2026 announcement, AI Mode has one billion monthly users. AI Overviews, the broader answer layer, reach 2.5 billion monthly active users across more than 200 countries.

Why are page-one rankings no longer enough for brand visibility?

In Ahrefs' analysis of Google AI Overviews, about 38% of cited pages also ranked in Google's top 10 for the same query. AI systems select sources based on freshness, entity authority, structured data, and third-party validation, not organic position alone. A number-one ranking that the AI engine does not cite is invisible to the majority of searchers who now consume answers directly.