Alphabet’s AI showcase is its chance to wow Wall Street
An Android character is displayed in front of a building on the Google headquarters campus in Mountain View, California, on July 23, 2025.
Justin Sullivan | Getty Images
AlphabetIts shares are up 140% in the past year as its cloud business grows faster than ever before. Amazon’s And Microsoft’s.
But 18 months ago, it looked like Google’s parent company had spent a decade preparing for the age of artificial intelligence, only to watch OpenAI define the market.
Wall Street now values Alphabet as one of the few companies positioned to profit from every layer of the prolific AI boom.
Google I/O, which starts on Tuesday, has always been a venue to show developers where the company is going. The risks are higher this year.
Wall Street has already rewarded Alphabet for its AI comeback, but investors want to see whether that confidence is backed by a real product roadmap in key areas like search, cloud, Android, chips and enterprise software.
“Google is probably the company best positioned to monetize AI at scale because AI controls almost every layer of the stack,” said Lo Toney, founding managing partner of Plexo Capital and an early investor in Anthropic. “We have never seen a company that has complete vertical integration from top to bottom to be able to support AI.”
Gene Munster, managing partner of Deepwater Asset Management, said the advantage of having so many layers of control is not just scale but also speed.
“There is a benefit to having the entire stack in terms of the speed at which you can innovate,” Munster said. “For example, when you build your own custom silicon, that’s a speed advantage. Once you have access to power, you can get data centers up and running faster. That’s a speed advantage, and that’s important.”
Here are seven key areas investors are watching at Google I/O:
What’s next for the twins?
The most followed announcement will be whether Google will introduce the new generation Gemini model.
While reports ahead of I/O pointed to a potential launch of Gemini 4, analysts aren’t fully confident in it. Citi notes that Google is on a launch cadence of about three to four months, with Gemini 3.1 Pro released in February, and that a Gemini 3.2 or 3.5 update is more likely than a full generation jump.
This makes the Gemini 4 question more than a version number. A full generation leap would give Google a clearer response to OpenAI and Anthropic. Mizuho wrote that the Gemini 4 announcement “will push Google to the frontier,” while another 3rd-gen update will be more of a catch-up.
The broader Gemini ecosystem update will also be important.
Mizuho analysts said they will specifically monitor progress on Project Astra, Google’s universal AI assistant, as well as deeper Gemini Live features in Search, Gmail, Calendar and Maps, screen sharing, video understanding and native widget use. Updates are also expected for Google’s open source model family Gemma and Gemini Robotics.
Usage numbers entering the event are already stronger than they were a year ago. Paid Gemini Enterprise’s monthly active users increased by 40% in the first quarter compared to the previous quarter. According to Citi data, the number of monthly active users of the Gemini app in the US increased by 127% in April compared to the previous year. Token consumption reached 16 billion per minute, starting with Google Cloud Next.
artificial intelligence agents
If there’s one theme that runs throughout the I/O session sequence, it’s brokers.
Google has sessions on mediated coding workflows, multi-modal tools, media creation, robotics, and AI agents. The goal is to position Gemini as more than a chatbot and as an operational layer that can understand context and take action across Google products.
“He’s the one winning the office co-pilot market,” Tony said. “If the larger market becomes and drives AI agents – inference infrastructure, multi-modal workflows, enterprise search – that’s where we see a huge opportunity for Google to power Alphabet’s future growth.”
Agency coding is part of positioning Gemini as the answer to Anthropic’s Claude Code and OpenAI’s Codex. This category has become one of the clearest demonstrations of the business value of AI, especially in enterprise software.
agency shopping
Trading may be a bigger opportunity. Google already has search, shopping, autofill, and payments; now they want Gemini to connect them to a mediated payment experience.
Google is expanding its Universal Commerce Protocol, adding partners including: MetaMicrosoft, Ribbon, Klarna And Confirm in recent weeks. I/O is expected to demonstrate how this infrastructure can enable end-to-end mediated payment, where Gemini not only answers a shopping query but also completes a transaction.
Google Android Ecosystem Head Sameer Samat explained that he asked Gemini to plan a barbecue, create a menu, open Instacart, add items to the Safeway cart, and notify him when the task is completed.
“If you add that up multiple times a day throughout the week, that goes back a lot of time,” Samat said. “These are features that people are much more excited about and I think are much more tangible.”
Toney said Google’s multi-modal experience gives it a structural advantage as workflows become more complex.
“Enterprise workflows increasingly include things like video, audio, images and code,” he said. “Google is uniquely strong in multi-modal systems because it has that experience with some of the largest applications that drive them (YouTube, Android, Maps, Search, DeepMind) and of course TPUs.”
For investors, the agency trading move has implications beyond Alphabet. Mizuho noted that Google’s more effective product development could focus on markets such as: Reservation Holdings, Expedia, Door Indicator, Zillow And InstacartHe noted that this expectation of change is likely part of the recent weakness in these stocks.
Artificial Intelligence Mode
The next question is how Google gets paid for this. AI-powered campaigns now account for more than 30% of search spend, according to Citi. AI Max, which came out of beta in April and will replace Dynamic Search Ads in September, is showing early results, including stronger conversions, with its full-featured suite.
Citi noted that AI Mode could help Google monetize longer, more complex queries that have historically been harder to convert into ad money. But Mizuho points out a paradox: AI Mode searches generate far fewer outbound clicks; The company estimates that 93% of AI Overview queries end without an external click, and organic click-through rates are down 15%.
This makes monetization one of the biggest questions for I/O. Munster said he’ll be watching for new ad products within AI Mode, how Google frames agent commerce, and what it says about more personalized AI experiences.
Google Cloud
But the most important I/O announcements for investors may come from cloud and infrastructure.
Cloud has become one of Alphabet’s strongest pillars. It grew by 63% in the first quarter compared to the same period last year, surpassing both Azure and AWS. Cloud deposition volume reached $462 billion, up roughly 90% from the previous quarter, with half expected to be realized within the next 24 months. Generation AI product revenue increased nearly 800% year-over-year.
CEO Sundar Pichai noted several factors behind this growth, including faster new customer acquisition, larger commitments and deeper relationships with existing customers.
The number of billion-dollar-plus deals signed in 2025 exceeded the previous three years combined, while existing customers exceeded their initial commitments by more than 30%, the company said in its first-quarter earnings call.
AI chips
The new wild card is external TPU sales. Google announced in the first quarter that it would begin offering its custom AI chips to external customers in the second half of 2026, with a broader expansion planned for 2027. This is a potentially huge new revenue stream, but investors still don’t quite know how to model it.
Investors will be listening for every detail about whether external TPU sales are recorded as gross sales or royalty income, what margins look like and how those deals are accounted for in backlogs, Mizuho wrote.
Toney called TPUs one of the most underappreciated parts of Alphabet’s thesis, arguing that Google’s in-house chips allow the company to build a tightly integrated AI infrastructure that supports not only Gemini and Cloud, but also YouTube, Android, and the rest of its ecosystem.
Munster estimates the broader AI chip market is running at around $500 billion annually, meaning even modest share gains could be significant for Alphabet.
anthropic
No relationship attracts more scrutiny regarding I/O. Google’s ties to Anthropic.
Alphabet has a significant ownership stake in the AI startup, and its recently reported $200 billion cloud commitment, if true, could represent a large chunk of Google’s contracted future cloud revenue.
Google has also committed to investing up to $40 billion total, creating a loop in which capital flows to Anthropic and back to Google through cloud and TPU spending.
This dynamic raises an issue of concentration in the cloud that investors have already seen elsewhere. SeerIts shares rose after reporting a huge increase in backlogs, largely due to OpenAI, then sold off as investors became more nervous about customer concentration. Microsoft is facing a similar controversy regarding its OpenAI relationship.
Toney said the Google-Anthropic relationship is more like a hedge than a weakness.
Even if businesses choose Claude over Gemini, Google can still benefit from the demand for the infrastructure behind that usage, he said.
“If businesses choose Claude, then Google still wins in infrastructure because all of that activity has to happen somewhere,” he said. “Google still wins because of its TPUs.”
WRISTWATCH: Inside Alphabet’s return on capital spending: What investors need to know
