Alphabet Plans Its First Stock Sale in 20 Years to Fund AI, With $10B From Berkshire

On June 1, 2026, Alphabet announced plans to raise roughly $80 billion by selling new stock, the company's first equity offering since 2005. The cash is earmarked almost entirely for one thing: building more AI computing capacity. For a company that funded two decades of growth from its own profits, the decision to tap public markets is a striking signal of how expensive the AI race has become.

Key takeaways

• Alphabet announced an $80 billion equity raise on June 1, 2026, later upsized to about $84.75 billion as demand poured in.
• The package has three parts: $15 billion in mandatory convertible preferred stock, $15 billion in Class A and Class C shares, and a $40 billion at-the-market program starting in Q3.
• Berkshire Hathaway bought $10 billion directly in a private placement, split between Class A shares at $351.81 and Class C at $348.20.
• Alphabet expects 2026 capital expenditures of $180 to $190 billion, roughly double 2025, with 2027 set to rise again.
• It is Alphabet's first stock sale since its 2005 follow-on, underscoring the scale of the AI infrastructure buildout.

What happened

Alphabet, the parent of Google and YouTube, said it would sell about $80 billion in new shares and related securities. The package has three parts: $15 billion in mandatory convertible preferred stock, $15 billion in regular Class A and Class C shares, and a $40 billion "at-the-market" program that lets the company sell shares gradually starting in the third quarter.

The offering was later upsized to roughly $84.75 billion as demand came in. Alphabet also agreed to a private placement that sold $10 billion of stock directly to Warren Buffett's Berkshire Hathaway, split between $5 billion of Class A shares at $351.81 each and $5 billion of Class C shares at $348.20 each. The placement added to a position Berkshire had been quietly building since the third quarter of 2025.

The company said the proceeds are for general corporate purposes, "including capital expenditures to scale AI infrastructure and global compute."

Here is how the roughly $84.75 billion package breaks down:

Component	Size	Detail
Mandatory convertible preferred	$15 billion	Converts to common stock on a set future date
Class A and Class C shares	$15 billion	Sold in the underwritten offering
At-the-market program	$40 billion	Shares sold gradually from Q3 2026
Berkshire private placement	$10 billion	Class A at $351.81, Class C at $348.20
Upsize from added demand	~$4.75 billion	Took the total above the initial $80 billion

The details

The numbers behind the raise tell the story. On its first-quarter 2026 earnings call, Alphabet guided to capital expenditures of $180 to $190 billion for the year, roughly double its 2025 figure, and said 2027 spending would increase significantly again. Most of that goes to data centers packed with AI accelerators, the power and cooling to run them, and the networking to tie them together.

Even a company sitting on enormous cash reserves apparently decided that tapping public markets was the cleaner way to fund a buildout of this scale. The appetite for high-bandwidth memory and AI accelerators driving these costs is the same force lifting suppliers across the industry, a dynamic visible in Micron's record AI-memory quarter.

Why it matters

For two decades, Alphabet funded itself entirely from its own profits and never needed outside equity. Choosing to sell stock now, even diluting existing shareholders slightly, signals just how capital-intensive the AI era has become. The Berkshire investment adds a notable endorsement: Buffett's firm has historically been cautious about fast-moving technology bets, so a $10 billion vote of confidence carries symbolic weight beyond the dollars.

Alphabet is not alone. Across the industry, the largest technology companies are pouring record sums into AI infrastructure, and several have turned to debt or equity markets rather than relying purely on operating cash flow. The same pressure is reshaping workforces, too, as seen when Oracle cut 21,000 jobs while pointing partly to AI and redirected spending toward data centers.

The bigger picture

The risk is straightforward: if AI demand grows more slowly than these capital plans assume, companies will have spent hundreds of billions on hardware that depreciates quickly. Chips and servers do not stay state of the art for long. That tension, huge upfront spending against uncertain long-term returns, is the central debate hanging over the entire AI buildout.

There is also a counter-pressure inside many enterprises to wring more value from existing AI spend rather than expand endlessly, a shift we explore in why companies are suddenly counting every AI token. Whether that frugality dents the demand justifying Alphabet's raise is one of the open questions.

What is next

A few things to watch:

• How fast the $40 billion at-the-market program runs. The pace of those sales will hint at how aggressively Alphabet wants to deploy capital.
• Capex guidance for 2027. Management has signaled spending may rise again, so the next earnings updates will be closely read.
• Whether other giants follow. If more large technology firms raise outside capital for AI, it confirms that internal cash flow alone can no longer keep up.
• Returns on the spending. Investors will eventually want evidence that all this compute is translating into durable revenue, not just capacity.

Frequently asked questions

Why does Alphabet need outside money if it is so profitable?

Even with large cash reserves, the scale of AI infrastructure spending, $180 billion-plus in 2026 alone, is enormous. Raising equity lets Alphabet preserve its cash position and balance sheet flexibility while still funding the buildout.

Will this dilute existing shareholders?

Yes, modestly. Issuing new shares spreads ownership across a larger base. The at-the-market structure is designed to minimize price impact by selling gradually rather than all at once.

Why is Berkshire's investment notable?

Berkshire Hathaway has historically avoided fast-moving tech bets, so a $10 billion private placement reads as a strong endorsement of Alphabet's AI strategy and long-term cash generation.

What is the main risk to this strategy?

Overbuilding. If AI demand grows slower than planned, Alphabet could be left with expensive, fast-depreciating hardware. The payoff depends on AI revenue scaling to match the spending.

For now, the message from Alphabet is clear. The company that went 20 years without selling a share has decided the AI era is worth breaking that streak.