Google’s free cash flow for the second quarter of 2026 turned negative for the first time in the tech giant’s history as a publicly traded company, driven by record spending on artificial intelligence.
Google ended the quarter with a negative free cash flow of $5.9 billion, company executives shared in the earnings call, and updated expectations to spend up to $205 billion in 2026, up from previous guidance of $180 billion to $190 billion. The capital expenditures will also “increase significantly in 2027,” parent company Alphabet’s chief financial officer Anat Ashkenazi said on the call. Analysts expect that number to at least hit $262 billion, according to Bloomberg.
The reason? Usual suspect: the AI investment boom.
“We expect that free cash flow will remain under pressure driven by our investments in technical infrastructure, which enable us to capitalize on the AI opportunity and continue to drive attractive returns,” Ashkenazi said.
The tech industry is pouring trillions of dollars into building out supply for what they claim is unprecedented AI demand. But some experts fear the investment is simply too big and unwarranted, raising fears of a potential AI bubble. The free cash flow of the four AI hyperscalers -Google, Meta, Microsoft and Amazon- was expected to go negative by 2027 after the eye-watering capex commitments announced in the last round of earnings. Some investors are worried that Silicon Valley might be recklessly burning through cash in its AI buildout goal with limited returns. If the returns on that investment don’t pan out as expected, it’s bound to hit Nvidia’s financials as well, considering that the chipmaker counts all four hyperscalers as major customers.
That’s why Nvidia CEO Jensen Huang has spent the last couple of months trying to ease Wall Street worries over this scenario, going on to say in the company’s earnings call in March that he is “confident” that the hyperscalers’ cash flow will grow, because “we have now seen the inflection of agentic AI and the usefulness of agents across the world in enterprises everywhere.”
That fear is also one of the likely reasons why Apple, which has largely steered clear from the AI spending gold rush at the expense of lagging peers, has outperformed its hyperscaler peers in the stock market and even briefly dethroned Nvidia as the most valuable company last week.
To quell worries at least a bit for Google fanatics, the company did provide numbers showing a pretty robust cloud business. Cloud sales hit $24.77 billion for the quarter, up a whopping 82% from last year.
“We are seeing very strong demand both from external cloud customers as well as across the business, and our goal is to invest as long as we see an attractive return on that investment,” Ashkenazi said.
Despite the success on the cloud side and the 24% overall revenue growth, sales came in below investor expectations for the company’s bread and butter, Search. That’s even though Google CEO Pichai said Search usage hit an all-time high this summer thanks to the World Cup.
Also putting pressure on Google is its performance in the AI products scene.
Google burst onto the frontier AI models scene with Gemini 3 and its image generator Nano Banana Pro last year, causing much of the internet to question whether OpenAI’s leading position in the industry was officially under threat from the tech giant. But since the stellar debuts in November 2025, the company has been rather silent as competitors like Anthropic boast flashy debuts one after another.
Google did, however, unveil Gemini 3.6 Flash earlier this week, but the model appears to lag behind the latest releases from OpenAI and Anthropic in most major benchmarks, and even behind Musk’s Grok 4.5 on some tests. Meanwhile, the promised Gemini 3.5 Pro, its most powerful flagship model, is reportedly months behind schedule. On Wednesday’s earnings call, Pichai said the Gemini 3.5 Pro model was currently in testing, and also teased Gemini 4, promising an ambitious new model.
“We are now training Gemini 4, and we’re being very ambitious with it,” Pichai said. “We wanted to compete at the frontier level of where the frontier will be when Gemini 4 comes out, and so we are applying a lot of our compute and effort in that direction,” Pichai said. “But with that, we are creating a baseline on top of which you will see us rapidly iterate with subsequent model releases, and so picking up pace and releasing models almost at a monthly cadence is part of our roadmap as we are building Gemini 4.”
English (US) ·