This article was AI-generated as part of an experimental historical-content project. The date reflects the period being analyzed rather than the date the article was originally written.
Alphabet’s second-quarter call had the number everyone expected. In his prepared remarks, Sundar Pichai said that since AI Mode expanded globally last October, “we’ve surpassed one billion monthly active users.” Google had already used a similar figure at I/O in May. Today it went into the investor record.
The rest of the quarter supported the story. Revenue was $119.8 billion, up 24%. According to the call transcript, Search & Other revenue grew 17% to $63.3 billion, with retail and finance making the largest contributions. Pichai said AI features in Search now send “billions of clicks to websites every week” and that the cost of AI Mode responses had fallen to its lowest level since launch. Full-year capital expenditure guidance went up to between $195 billion and $205 billion.
For investors, the message is that generative Search is no longer an experiment. It is scale, it is getting cheaper to serve, and it is being monetized.
What the call measured, and what it did not
Philipp Schindler described the ad formats being tested inside AI Mode. They include contextual sitelinks added to text ads, Direct Offers, and Highlighted Answers, which he described as “clearly marked sponsored links inside list responses.”
That last one is worth pausing on. In a conventional results page, ads and organic results sit in visibly separate places. In a list answer, a sponsored entry sits in the same structure as the generated recommendations. Google says it is clearly marked, and I have no reason to doubt that. But it means the answer itself is now a commercial surface, and the companies named in it, paid or not, are being presented side by side.
Nothing in the prepared remarks or the analyst questions measured whether AI Mode’s answers about companies are accurate. That is not a criticism of the call. Earnings calls report what investors price: users, revenue, cost, capital. Accuracy is not on that list, and there is no standard way to report it.
Why this belongs on the IR agenda
The gap matters most to the companies being described.
A billion monthly users means AI Mode is now one of the largest places where a company gets summarized. Some of those users are investors, analysts and journalists. When one of them asks AI Mode about a listed company’s leadership, a past controversy or its competitive position, the answer arrives in a format designed to feel complete.
Investor relations teams already watch how their story appears in filings, news and analyst notes. The results page has been part of that picture for years. I argued in 2023 that markets move on filings, but reputation hardens on the results page. Generative answers are now a large part of that results page, and they summarize with more confidence than a list of links ever did.
So a reasonable addition to an IR and communications calendar is simple: after each earnings release, check how AI Mode and AI Overviews describe the company, its numbers and its management, and which sources they cite. Do it again after a major announcement. Record the gaps.
The asymmetry
Google’s incentives are clear and public. Usage, monetization and efficiency are reported every quarter, and the capital commitment just went up again. The accuracy of what those billion users are told about any specific company is not reported anywhere.
That is not Google’s job to publish. It is the job of the companies being described to notice.