Alphabet and Microsoft: AI-driven cloud accelerates, and so do investments
On October 28, Google Cloud doubled its year-over-year revenue (+103%, $30.7 billion) and Azure grew 48%. Meanwhile, Alphabet spent $51.4 billion in capital expenditures for the quarter and Microsoft $39.4 billion.
Alphabet and Microsoft released their results for the quarter ended September 30 on the same evening, October 28. Both say the same thing with different numbers: demand for AI compute is exploding their cloud businesses, but it comes at a price — investments that have more than doubled in a year.
Google Cloud doubles, Alphabet borrows and raises capital
Alphabet’s revenue reached $127.6 billion (+25%). The surprise came from cloud: Google Cloud jumped to $30.7 billion (+103%), versus $15.2 billion a year earlier, and its operating income nearly tripled, to $11.1 billion (versus $3.6 billion). The release cites infrastructure and enterprise AI solutions, as well as sales of TPU systems, Google’s in-house chips, which “contributed significantly” to growth.
On usage, Sundar Pichai stated, without methodological detail, that the Gemini app has reached one billion users and that Google’s model APIs process nearly 32 billion tokens per minute, more than four times as much as a year ago.
The flip side appears in the cash flow statement. Alphabet spent $51.4 billion on capital expenditures for the quarter, versus $24.0 billion a year earlier, against $42.3 billion in cash generated from operations: free cash flow was therefore negative by roughly $9 billion. Alphabet bought back no shares this quarter ($11.5 billion a year earlier), and raised $12.4 billion in common stock and $29.0 billion in debt. Net income fell to $26.2 billion, weighed down by an $8.7 billion loss on equity investments.
Azure up 48%, a $677 billion order backlog
At Microsoft, revenue reached $92.3 billion (+19%). Microsoft Cloud came in at $63.9 billion (+30%) and Azure at $33.2 billion (+48%), versus $22.4 billion a year earlier. The commercial remaining performance obligation — contracts signed but not yet billed — climbed to $677 billion (+73%). The new “Agents and Infra” segment posted $76.8 billion (+25%), while Windows OEM and devices declined 22%.
Investments followed the same curve as at Google: $39.4 billion in capital expenditures for the quarter, versus $19.4 billion a year earlier.
What the guidance says
During the analyst call, Alphabet CFO Anat Ashkenazi maintained the investment range of $195 to $205 billion for 2026 and announced a “significant” increase in 2027, without a figure, deferred to next quarter. She also indicated that Google Cloud is short on capacity: the group began renting compute from third parties in late September, which will weigh on margins in the near term. Neither Alphabet nor Microsoft publishes the exact share of AI in their cloud growth.
Our take
These results partly answer the question weighing on the sector: is AI already paying off? For cloud, yes, and fast: a Google Cloud that doubled in a year and Azure at +48% show that enterprises are paying for compute and models. But spending is growing at the same pace. Alphabet is now financing its data centers partly through debt and stock issuance rather than its own cash alone, and has suspended its share buybacks, a sharp change for a group long known for exceptional profitability. Microsoft’s $677 billion backlog provides visibility for years to come. And when Alphabet admits it lacks capacity to the point of renting compute from third parties, it confirms that compute supply still isn’t keeping up with demand, despite these record expenditures.
Sources
- Alphabet — Q3 2026 earnings release (Exhibit 99.1, SEC) Primary source
- Google — remarks by Sundar Pichai at the Q3 2026 earnings call Primary source
- Microsoft — fiscal year 2027 first quarter earnings release Primary source
- Microsoft — earnings release (Exhibit 99.1, SEC) Primary source
- The Motley Fool — Alphabet analyst call transcript Secondary source — Third-party transcript; used only for investment forecasts.
A primary source comes directly from the company, lab or organization concerned. A secondary source reports the information without being its author: we flag it as such. Spotted an error? See our corrections policy.
