Google's AI Mode has surpassed 1 billion monthly users, the company says, matching ChatGPT's estimated monthly active user base as of May, according to Sensor Tower data. The figure marks the clearest confirmation yet that Google's strategy of embedding AI chat directly into its dominant search engine has converted its existing audience at scale — not by attracting users who chose AI, but by routing them into it.

The mechanics matter for anyone tracking search behavior. Users who ask certain questions now see an AI Overview, followed by a search bar encouraging them to "Ask Anything." Clicking it drops them into a chatbot conversation inside AI Mode. Millions of people who never intended to use a chatbot have acclimated to one through this pathway. In recent months, Google has made the funnel more aggressive: the "Ask Anything" bar is often no longer hidden behind a "Show More" button, and many users now see an AI Mode button directly on the Google Search homepage.

The numbers behind the shift

Google Search usage hit an all-time high this year, and much of that activity involves AI. Confirmed, tool-based signals fill in the picture:

• AI Overviews now appear in 40% of U.S. Google searches, according to Similarweb's 2026 Generative AI Landscape report.
• A Pew poll this summer found that 60% of Americans read AI summaries at the top of search results, versus 49% who use chatbots — meaning more Americans encounter AI through Google's interface than through standalone AI products.
• Queries inside AI Mode run three times the length of traditional search queries, according to Kevin Indig, a strategy advisor who writes the Growth Memo newsletter.

Indig draws a sharp behavioral distinction between the two surfaces. AI Overviews, introduced in 2024, did not drastically change how people used Google: users still scrolled past the summaries and clicked links to verify information. AI Mode is different. Users brainstorm essay ideas, build study guides, plan trips, and seek financial advice — and they largely stay inside the interface rather than exploring the web.

"It's crazy: they will get an answer, and they'll barely go out of their way to see if it's correct," Indig says. "That means very few clicks."

That trust gap is the core problem for publishers and retailers.

The traffic and revenue impact

Publishing is taking the hardest hit. Similarweb data analyzed by Axios shows that U.S. web traffic to roughly 100 of the top news sites fell from above 70 million visits in 2024 — the year Google introduced AI Overviews — to 47.6 million visits this July. The decline cannot be attributed entirely to AI, but a separate Ahrefs study found that AI Overviews were associated with a 58% reduction in clickthrough rate to top-ranking search results.

E-commerce faces a parallel disruption. The percentage of people who use AI to shop online has doubled in less than two years, per Similarweb — a shift that leaves retailers with far less first-party information about their customers. "You lose the direct relationship with customers," Indig says.

Even inside AI companies, the concern is documented. Newly unsealed court documents from The New York Times' copyright infringement lawsuit against OpenAI and Microsoft, made public earlier this month, included worries from both companies that their AI content strategies had started a "doom loop" that would "hurt the performance of our models and the entire web at the same time," and present an "existential threat" to publishers. That is an internal admission from Google's competitors, not a Google statement — but it describes the same economic dynamic Google's AI Mode accelerates.

Google's own position, from Robby Stein, vice president of product for Google Search: "It is critical to our vision to remain the connective tissue to the web. Users want to hear from experts. They want to be able to verify and feel good about the content they're seeing." Stein also says Google has facilitated a "dramatic increase in the amount of links in the content trays we're providing," and that a "thriving, healthy ecosystem of publishers and creators is of paramount importance."

Why distribution won

At the start of the generative AI race, Google was criticized for lagging behind OpenAI and Anthropic. Four years in, its advantage looks structural. Indig compares the dynamic to Snapchat versus Instagram: Snapchat invented disappearing stories, but Instagram copied the feature into an app with a larger installed base and kept its dominance. Google holds Instagram's position here.

"Younger generations prefer ChatGPT. But when people do research on ChatGPT for products, they will go to Google and look for the brand," Indig says. "So Google still has an important utility."

Gene Munster, managing partner at Deepwater Asset Management, argues the new search behavior could be habit-forming. "People need to have something 10 times better to really switch behavior," he told TIME in April. "And they have something that's actually really pretty good, but it doesn't need to be the best."

The strategy traces to CEO Sundar Pichai, who declared in 2016 that Google would be an "AI-first company." In April, Pichai told TIME: "I view AI as the most profound way we can make progress against that mission" of organizing the world's information.

Notably, Stein says AI Mode is reaching users who never would have created an OpenAI account: for some older users, "this is the first time they've experienced AI."

What to monitor

For publishers, retailers, and SEO practitioners, the metrics to watch are the link volumes inside AI Mode's content trays that Stein cites, the trajectory of news-site traffic after its drop to 47.6 million monthly visits, and whether Google's "connective tissue" rhetoric translates into measurable clickthrough recovery. The open question — one the source leaves unanswered — is whether the content creators Google depends on for quality information can withstand the traffic shifts its own interfaces have set in motion.