Alphabet’s cloud business grew 82% last quarter, yet the company spent more on infrastructure than it generated in operating cash. Free cash flow turned negative, share repurchases stopped and nearly $70 billion of new equity and debt entered the balance sheet.
None of that came from weak operating performance. Alphabet’s revenue increased 24% to $119.8 billion, operating income rose 30% to $40.8 billion and consolidated operating margin expanded from 32% to 34%. Google Cloud generated $24.8 billion of revenue, while its operating income increased from $2.8 billion to $8.8 billion.
Customers are buying more AI infrastructure, AI solutions and core cloud services than Alphabet can currently provide. Management said demand continues to exceed the capacity added during the past three years, and the company raised its 2026 capital-spending forecast again, from an original $180 billion to a new range of $195 billion to $205 billion. Another significant increase is already planned for 2027.
Building that capacity required $44.9 billion of property and equipment spending during the quarter, twice as much as one year earlier. Alphabet generated $39.1 billion of operating cash flow, leaving free cash flow at negative $5.9 billion.
The company also made no share repurchases after buying back $13.2 billion of stock during the same quarter last year. It raised $49.6 billion through common and mandatory convertible preferred equity and another $20.3 billion through senior unsecured notes, while long-term debt increased from $46.5 billion at the end of 2025 to $98.2 billion.
Alphabet didn’t say that all of the equity proceeds would be spent on AI. They were designated for general corporate purposes, including AI infrastructure and global compute. But the company raised that capital while infrastructure spending exceeded operating cash flow, repurchases went to zero and the spending forecast moved higher again.
For years, Alphabet generated enough cash to invest heavily, repurchase shares and maintain enormous balance-sheet flexibility without having to choose among them. The AI buildout is now large enough that those claims on its cash are beginning to compete.
The companies supplying the buildout don’t have to wait for Alphabet to earn a full return on the capacity. Chipmakers, memory manufacturers, server vendors, power-equipment suppliers and construction firms are paid as equipment is delivered and data centers are built. Alphabet commits the capital first and earns it back as customers consume more computing services over time.
That leaves shareholders giving up repurchases and accepting dilution, while creditors provide more capital before utilization and returns are fully known. The demand may be real, but Alphabet still carries the risk that capacity arrives faster than customers can absorb it or that the returns take longer to emerge than the spending cycle assumes.
The first market reaction reflected that tension. Alphabet shares fell about 3% in extended trading after the higher capital-spending forecast, while South Korea’s KOSPI rose more than 3% as SK Hynix and Samsung Electronics benefited from expectations of continued memory and infrastructure demand.
The companies receiving the orders were rewarded. The company financing them was asked to show that revenue and cash generation can eventually outrun the cost of building the next round of capacity.
Alphabet has plenty of room to make that case. It ended June with approximately $242.5 billion in cash, cash equivalents and marketable securities, and trailing 12-month free cash flow remained positive at $53.3 billion. Cloud operating income more than tripled, and the company raised capital from a position of strength rather than distress.
The issuance may have been opportunistic, and one negative quarter doesn’t mean Alphabet can no longer fund AI internally. If Cloud revenue and profit continue growing at their current pace, free cash flow could recover even while capital spending remains near the new forecast.
The next two quarters will show whether that happens. A strong return to positive free cash flow, followed by the resumption of substantial repurchases, would make this quarter look like a temporary mismatch between construction spending and customer consumption.
Microsoft, Meta and Amazon will help determine whether the financing change extends beyond Alphabet. If they can support comparable infrastructure programs without issuing more capital or reducing shareholder distributions, Alphabet’s experience will remain its own. If they can’t, the AI buildout will be changing how the largest technology companies finance growth before it becomes clear how much of that growth the new capacity can ultimately produce.