Alphabet Raises $80B in Stock to Fund AI Compute — Berkshire's $10B Private Bet Signals New Investor Class
Alphabet announced Monday it will raise $80 billion through a package of stock sales, the largest equity capital raise in the company’s history, to fund AI compute infrastructure it says is already supply-constrained.
The deal breaks into three parts: a $40 billion at-the-market program where Alphabet will sell shares from time to time starting in Q3; $30 billion in underwritten offerings of shares and mandatory convertible preferred stock; and a $10 billion private placement with Berkshire Hathaway — $5 billion in Class A shares at $351.81 each and another $5 billion in Class C stock at $348.20.
The Berkshire transaction is notable. Greg Abel’s first major capital allocation move as Berkshire’s CEO is a $10 billion direct check into the AI infrastructure thesis, deepening a position the conglomerate began building only three quarters ago. Berkshire has now turned Alphabet into one of its largest equity holdings, deploying a meaningful portion of its nearly $400 billion cash pile into a single AI infrastructure bet.
Alphabet was blunt about why: “The company is experiencing strong demand for its AI solutions and services from enterprises and consumers, at levels that are exceeding the company’s available supply.”
The Numbers Behind the Raise
Alphabet already revised full-year capital expenditure guidance upward in April, from a $175-185B range to $180-190B. The $80B raise closes a significant portion of the gap between operating cash flow and the compute buildout required.
Google Cloud posted $20 billion in Q1 2026 revenue, a quarter that Sundar Pichai described as “capacity-constrained.” That language — revenue bounded by what Alphabet could provision, not what customers wanted to spend — is the same logic that produced the raise. The company is not raising to speculate; it is raising because demand already exceeds what infrastructure can serve.
What It Goes Into
Alphabet says proceeds fund “capital expenditures to expand AI infrastructure and global computing capacity.” Concretely, that means TPU clusters, data center construction, and the networking and power infrastructure that runs them.
Google split its TPU line in two earlier this year — TPU 8t for training, TPU 8i for inference — and secured Marvell as an additional chip design partner. Google also signed a classified Pentagon AI deal and a $750 million commitment to accelerate agentic AI across its 120,000-partner ecosystem. All of it demands more physical compute.
The $80B raise, compounded with Anthropic’s $200 billion five-year cloud commitment to Google and Meta’s $21 billion CoreWeave/Vera Rubin deal, points to a 2026 where the binding constraint on AI progress is measured in megawatts and GPU-months, not model quality.
The Market Context
Alphabet’s raise lands at a moment when all three AI-adjacent IPO candidates — Alphabet itself (already public), Anthropic, OpenAI, and SpaceX — are testing how much capital markets can absorb. The Economist’s framing this week — can markets swallow Anthropic, SpaceX, and OpenAI simultaneously — frames the question around private-to-public transitions. Alphabet’s raise reframes it: even fully public giants need extraordinary capital infusions just to keep pace with AI demand.
Nine cloud giants are projected to spend $830 billion on AI infrastructure in 2026. Alphabet’s $80B raise is its share of the arms race, written in equity rather than debt.