Berkshire Hathaway Anchors Alphabet's AI Raise With a $10 Billion Private Placement

Berkshire Hathaway Anchors Alphabet’s AI Raise With a \$10 Billion Private Placement

When Alphabet needed an anchor for the largest equity raise in the history of publicly traded technology, it did not turn to a sovereign wealth fund or a syndicate of growth-chasing hedge funds. It turned to the most famously patient buyer in American finance. On June 1, Berkshire Hathaway agreed to plow \$10 billion into Google’s parent through a private placement — the cornerstone of an offering that priced two days later at a staggering \$84.75 billion. For a conglomerate that spent decades avoiding technology stocks on principle, it was a declaration: the AI capital-expenditure supercycle is now cheap enough, and certain enough, to warrant one of the biggest single tech checks Berkshire has ever written.

The mechanics are precise in the way Berkshire deals tend to be. The \$10 billion splits evenly: \$5 billion of Class A common stock at \$351.81 per share and \$5 billion of Class C capital stock at \$348.20 per share. It sits alongside \$30 billion in concurrent public offerings, a \$40 billion at-the-market program slated to run through the back half of the year, and a slug of mandatory convertible preferred — all of it earmarked, in Alphabet’s words, “to fund investments in its world-class AI compute infrastructure to meet its unprecedented customer demand.” The placement builds on a position Berkshire began accumulating in the third quarter of 2025 and expanded in early 2026. Post-deal, its Alphabet exposure ranks among the firm’s largest equity holdings, now the fifth-biggest in the portfolio.

Value investing meets AI capex

The through-line here is that Berkshire is not paying up for a story. Alphabet throws off roughly \$174 billion in trailing-twelve-month operating cash flow. Its Google Cloud backlog — contracted, not projected — sits near \$460 billion. Some 2 billion people a month touch a Gemini-powered product, and the physical estate underneath it all is difficult to replicate at any price: more than 10 million kilometers of fiber, 30-plus data centers, 40 cloud regions. In Buffett-Abel terms, this is not a bet on AI as a concept. It is a bet on a toll road with the compute already poured.

That distinction matters because the capex figures are the very thing spooking the rest of the market. Alphabet guided to \$180 billion to \$190 billion in capital spending for 2026, and the raise itself is what many investors read as confirmation that the AI buildout has grown too big to fund from cash flow alone. The stock has fallen by double digits since May. Berkshire looked at the same numbers and reached the opposite conclusion — that the spending is demand-driven, and that the market is discounting a franchise at a moment of manufactured fear.

The entry point

Timing is the tell. In the week of June 18–24, two marquee departures gutted sentiment. Noam Shazeer, a vice president of engineering and a central figure behind the Gemini models, left for OpenAI. Two days later John Jumper — the DeepMind engineering fellow who shared a Nobel Prize for AlphaFold — announced he was joining Anthropic. The exits triggered Alphabet’s worst trading day in more than a year and, across the broader selloff, wiped roughly \$269 billion from the company’s market capitalization. The private-placement price tags Alphabet at a clear discount to its February 2026 highs.

Buying into that kind of dislocation is the oldest move in the Berkshire playbook. “If you have a temperament that when others are fearful you’re going to get scared yourself,” Buffett told shareholders at the 2010 annual meeting, “you know, you are not going to make a lot of money in securities over time, in all probability.” He has been blunter still about price declines in the names he wants to own: “When it goes down, we love it, because we’ll buy more. And if it goes up, it kills us to buy more.”

Notably, this deal carries Greg Abel’s fingerprints, not just Buffett’s. Abel, who took over as CEO at the end of 2025, more than tripled Berkshire’s Alphabet stake and led the charge on the \$10 billion placement — the conglomerate’s first major AI investment out of a cash pile that has hovered near \$400 billion. “If he had any qualms about Alphabet spending more on AI infrastructure, he would never have committed such a significant amount of Berkshire’s capital,” wrote Motley Fool analyst Keith Speights, who called the pullback “an excellent buying opportunity for long-term investors.”

Is the talent loss a real risk?

The bear case is not frivolous. In frontier AI, a handful of researchers can define a model generation, and losing two at once — to your two most aggressive competitors — is not nothing. But the counterargument is structural. Alphabet retains one of the deepest benches of AI talent on earth, led by DeepMind chief Demis Hassabis, plus the balance sheet to recruit against any poaching campaign. CFO Anat Ashkenazi has framed the spending as a response to “unprecedented internal and external demand for AI compute resources” — the kind of demand that, if real, outlives any individual departure. Talent is mobile; 10 million kilometers of fiber and \$460 billion of contracted backlog are not.

There is also a signaling dimension that cuts against the fear trade. Alphabet joined the Dow Jones Industrial Average at the end of June, cementing its status as infrastructure to the U.S. economy rather than a speculative growth name. Berkshire arriving as an anchor buyer at a discount, in the same window, is the sort of validation that money-losing sentiment rarely survives for long.

What to watch

Three things. First, Berkshire’s next 13-F filing: if Abel used the June selloff to add beyond the placement, it confirms conviction rather than opportunism. Second, whether Alphabet’s cloud backlog keeps converting into recognized revenue at the pace that justifies \$190 billion of annual capex — the number that has to make the toll-road thesis real. Third, whether the talent exodus stops at two names or metastasizes into a broader migration toward OpenAI and Anthropic. For now, the most disciplined investor in the market has told you where he thinks the risk-reward sits. When Berkshire anchors a raise this size, the question is no longer whether AI infrastructure is worth building — it is who gets to own the tollbooth.