Google just posted the fastest growth its cloud business has ever recorded. Cloud revenue hit $24.8 billion in the second quarter of 2026, up 82% from the same period a year ago. To put that in perspective: a year ago that number was $13.6 billion. It has nearly doubled in four quarters.
This is not advertising money. Google's cloud unit earns revenue when companies pay to run their software, store their data, and use AI tools on Google's servers. The 82% growth came specifically from enterprise AI products and AI infrastructure, Google said. Nearly 90% of the Fortune 100 now use Gemini Enterprise, Google's AI product built for large organizations.
The backlog tells an even more important story. Google has $514 billion in signed cloud contracts that have not yet been billed. That figure was $240 billion just six months ago. It more than doubled in two quarters. When a company signs a multi-year cloud deal, it shows up in backlog first. The size and pace of that backlog growth signals that large organizations are not just testing AI tools; they are committing to them at scale.
The cloud unit's profit is also improving fast. Operating profit reached $8.8 billion this quarter, up from $2.8 billion a year earlier. That matters because it answers the core question investors have been asking: is this spending creating a real business, or just revenue with no margin? The answer right now is that margins are expanding, not shrinking.
Google also closed its $32 billion acquisition of Wiz, a cloud security company, earlier this year. That added security capabilities to Google Cloud and is contributing to its growth, particularly as companies moving sensitive data to the cloud demand stronger security guarantees.
The complication is that Google's own spending keeps climbing. The company raised its capital spending plan again on this earnings call, now guiding for $195 billion to $205 billion this year, up from $180 to $190 billion just last quarter. Its finance chief said demand still outpaces capacity, and 2027 spending will rise again. Analysts estimate 2027 spending could approach $260 billion or higher. Investors sent the stock down around 5% after the new figure was announced, even with the record revenue.
This is the central tension in the story. The revenue is real and growing fast. The contracts are signed. The profit margins are improving. But the spending required to keep up with demand keeps growing too, and the costs of today's data centers get spread across six years of accounting, meaning the drag on future profits is still building up in the background.
For any organization that uses or is considering cloud services, the practical read is straightforward. The market for AI-powered cloud tools has clearly crossed from experiment to standard operating budget for large companies. The vendor lock-in is real: once a company commits to a platform at scale, switching is costly. If your organization has not yet evaluated where your data and AI workloads will run over the next three to five years, the window for a relaxed, low-pressure decision is closing.