Microsoft, Google and Amazon Say AI Cloud Demand Still Outruns Supply

Week of October 26, 2026

Google Cloud Doubles as Cloud Giants Say AI Demand Outruns Supply

Microsoft, Alphabet and Amazon all reported faster cloud growth and said AI compute capacity, not customer interest, is the main limit.

View the data

	Revenue growth
Google Cloud	103%
Microsoft Azure	48%
AWS	39%

More charts in By the Numbers

Microsoft, Alphabet and Amazon all reported faster cloud growth this week, and all three said the same thing in different words: enterprise demand for AI compute is still bigger than the capacity they can bring online. For CIOs and chief AI officers, the message is that AI infrastructure is likely to stay scarce and expensive into 2027, even as the companies building it commit to spending more than ever.

The results also shook investors. Alphabet shares fell 13% on Thursday, October 29, the stock's worst single-day drop on record, after search and advertising sales came in below forecasts, The National reported. The next day the Nasdaq Composite lost 2.86% and a broad semiconductor index fell 7.2% after weak results from Intel and Amazon, Reuters reported.

What did Microsoft, Google and Amazon report?

Microsoft reported fiscal first-quarter revenue of $92.3 billion on October 28, up 19%. Azure brought in $33.17 billion and grew 48%, the first time Microsoft has disclosed Azure revenue in dollars, according to SiliconANGLE. Capital spending, including finance leases, reached $45.2 billion, below the roughly $50.7 billion analysts had expected. About three quarters of that now goes to short-lived assets such as GPUs and CPUs. Microsoft guided to Azure growth of 49% to 50% in constant currency for the December quarter. Chief Financial Officer Amy Hood told analysts, "We still have demand that exceeds supply."

Alphabet said Google Cloud revenue more than doubled, rising 103% to $30.7 billion, with operating income of $11.1 billion, in its quarterly earnings release filed the same day. Capital expenditures were $51.4 billion, up from $24.0 billion a year earlier. On the earnings call, executives put the cloud backlog at $561 billion, said the business remains supply-constrained and disclosed that Google began renting third-party compute late in the quarter. Google Cloud CEO Thomas Kurian said nearly 90% of the Fortune 100 use Gemini Enterprise. Alphabet kept its 2026 capital spending forecast at $195 billion to $205 billion and said spending will rise significantly in 2027.

Amazon followed on October 29. AWS revenue rose 39% to $46.0 billion, its fastest growth in nearly five years, with operating income of $17.8 billion, Amazon said. Free cash flow over the trailing twelve months was an outflow of $42.4 billion as spending on property and equipment climbed, mostly for AI infrastructure. Amazon still expects about $220 billion in 2026 capital spending, and CEO Andy Jassy indicated that 2027 spending will be meaningfully higher, CNBC reported, according to an AI Weekly summary. In a company post, Jassy put the AWS backlog at $578 billion and said Fortune 500 spending on Amazon Bedrock grew nearly sevenfold from a year earlier.

Why does the AI capacity crunch matter to enterprises?

Google Cloud and AWS alone now carry a combined backlog of more than $1.1 trillion, and all three providers described capacity, not customer interest, as the main limit on growth. That has direct consequences for enterprise buyers.

First, scarce compute tends to go to customers with large committed contracts. Companies planning to scale agents or inference-heavy workloads in 2027 should expect tougher talks over reserved capacity and should ask vendors how they allocate it.

Second, the buildout is increasingly financed with debt. Alphabet's free cash flow was negative in the quarter, at about $9.1 billion, and its long-term debt has risen from $46.5 billion at the end of 2025 to $125.1 billion, the Motley Fool noted. Providers spending at this pace will want returns, so AI prices may not fall as fast as many business cases assume.

Third, cost discipline is now a competitive skill. Speakers at AI Connect events have pushed this point for months. "You don't need premium models for everything that you're doing," Pushkar Rege, Vice President and Co-Founder of the Business Intelligence Program at Mastercard, said at The AI Enterprise Conference 2026, in a panel on tracking AI cost per outcome. Right-sizing models and measuring cost against business results matter more when capacity is tight.

How did investors and analysts react?

The market split the results by business, not by AI exposure. Microsoft shares were roughly flat on Thursday despite strong cloud numbers, while Alphabet's slide reflected concern that AI answers could erode its search business, according to The National. Michael O'Rourke of JonesTrading said Google was competing in AI "from a position of weakness." On Friday, Intel fell 28.6% after its own results, and server makers Dell and HP dropped about 11% and 10%, as investors questioned whether the spending would earn a return, Reuters reported.

For enterprise leaders, the useful signal is that no provider suggested demand is cooling. The open questions are how fast new data centers and power come online, and whether enterprises can show returns that justify the prices those buildouts require. Teams already making AI programs pay for themselves and controlling GPU and inference costs are best placed for that test.

What to watch next

Alphabet has said it will give 2027 spending details on its next quarterly call. Watch for new data center capacity coming online, changes in how providers price reserved AI capacity, and whether Microsoft's lighter capital spending this quarter holds.

Frequently asked questions

How fast did Azure, Google Cloud and AWS grow last quarter?

Azure grew 48%, Google Cloud revenue rose 103% to $30.7 billion and AWS grew 39% to $46.0 billion, according to the companies' late-October reports.

Why are cloud providers capacity-constrained?

Enterprise demand for AI compute is growing faster than providers can add data centers, chips and power, which executives at all three companies acknowledged.

What should enterprises do about AI capacity limits?

Negotiate reserved capacity early, ask vendors how they allocate scarce compute, and right-size models while tracking cost per business outcome.

Filed under: Data & AI Infrastructure, AI Strategy & ROI