Alphabet has upsized its equity raise to roughly $84.75 billion in gross proceeds, an extraordinary capital-markets move for a company that generates tens of billions in free cash flow every quarter. The structure is layered: about $30 billion in concurrent underwritten offerings -- split between $15 billion of mandatory convertible preferred depositary shares and $15 billion of Class A and Class C common -- plus a $40 billion at-the-market program expected to begin in the third quarter, and a $10 billion private placement to Warren Buffett's Berkshire Hathaway. The proceeds are earmarked for general corporate purposes, chiefly the capital expenditures needed to scale AI compute and global data-center capacity.
The headline isn't the dollar figure -- it's the decision to raise equity at all. Alphabet could fund enormous capex from operating cash, so choosing to issue stock, accept dilution, and bring in outside capital says the AI buildout has reached a scale where even Google wants to spread the financing risk. Management now guides 2026 capital expenditures to a range of $180 billion to $190 billion, with 2027 expected to increase significantly from there. That is a number that would have been unthinkable for any single company a few years ago.
“Management now guides 2026 capital expenditures to a range of $180 billion to $190 billion, with 2027 expected to increase significantly from there.”
The Berkshire placement is the detail that should make every skeptic pause. Warren Buffett built his reputation avoiding technology he didn't understand and refusing to pay up for narratives. A $10 billion private placement into Alphabet's AI-infrastructure raise is the value world's clearest endorsement yet that the compute buildout is a durable, cash-generative business rather than a bubble -- and it gives Alphabet a marquee anchor investor as it taps public markets.
For the rest of tech, the read-through is that the AI race has fully become a balance-sheet contest. When the most cash-rich company in the world raises equity to keep pace, the message to everyone below it is stark: access to capital and the willingness to spend it are now the binding constraints, not engineering talent. Hyperscalers will keep outbidding each other for chips, power, and land, and the cost of staying at the frontier is compounding.
What to watch next: how GOOGL trades through the at-the-market program, since continuous issuance is a slow, steady source of dilution that the market has so far waved through; whether Microsoft, Amazon, and Meta follow with their own capital raises; and whether the capex actually converts into cloud and AI revenue fast enough to justify the spend. For founders, the lesson is that the infrastructure layer is being financed at a scale no startup can match -- so build where capital intensity is a moat in your favor, not against you.