Google (GOOG) Shares: Why They Tumbled in June 2026

Google (GOOG) shares experienced their worst session in a year on 22 June 2026, falling nearly 7% to open at $356.83. It wasn't random panic. Behind the drop lies a rare combination — the departure of two central AI figures from Google and an infrastructure bill approaching $190 billion. This post organises what really moves the stock, separating signal from noise.

TL;DR

• Google (GOOG) shares fell around 7% on 22/06/2026, well off the all-time high of $404.47 set on 18/05/2026.
• The immediate trigger was human: Noam Shazeer, co-lead of Gemini, left for OpenAI, and John Jumper, 2024 Nobel Prize in Chemistry, left for Anthropic — on consecutive days.
• Alphabet raised its 2026 capex to $180–190 billion, six times the 2022 level and double the 2025 spend.
• In June, the company raised $84.75 billion in equity (above the initial $80 billion target), with $10 billion from Warren Buffett's Berkshire Hathaway.
• The counterpoint: Google Cloud grew 63% in Q1 2026, driven by AI. The long-term thesis remains intact.

Why Google shares fell in June 2026

The drop in Google shares didn't come from poor financial results. It came from a crisis of confidence combined with a bill that's growing faster than the revenue it promises to generate.

On 22 June 2026, GOOG stock retreated nearly 7% — its worst day in about twelve months. To put it in perspective: just over a month earlier, on 18 May, the stock had hit its all-time high of $404.47. The correction therefore erased a good chunk of a recent rally.

The market wasn't reacting to weak sales. It was reacting to two signals that together shake the narrative of Google's AI leadership: people leaving and money leaving. When both happen in the same week, investors start asking whether the billion-dollar return promised by AI will actually materialise — and when.

That's the central point of this article. The short-term volatility in Google shares says less about the current quarter and more about a long-term question: how much does it cost to lead the AI race, and who stays to do the work.

The talent exodus: what the departures of Shazeer and Jumper mean

The week's trigger was the loss of two heavyweights in a few days. In a field where the scarcest asset is not money or chips, but brains, that weighs heavily.

Noam Shazeer to OpenAI

Noam Shazeer was vice president of engineering and co-lead of the Gemini models. On 18 June 2026, he announced his move to OpenAI — the direct rival. The detail that made the headline sting: in September 2024, Google had paid around $2.7 billion to bring him back via the acquisition of startup Character.AI. Less than two years later, he crossed the street.

John Jumper to Anthropic

The next day, 19 June, John Jumper of Google DeepMind announced his departure to Anthropic. Jumper is no ordinary name: he won the 2024 Nobel Prize in Chemistry for AlphaFold, the AI system that solved the protein folding problem. He spent nearly nine years at DeepMind.

Losing a co-lead of Gemini and a Nobel laureate to the two biggest rivals in 48 hours is the kind of event the market reads as a credibility blow. It doesn't topple the technology overnight — Gemini remains standing — but it fuels doubt about Google's ability to retain the people building the future. And it's that perception that is reflected in Google shares.

There is also a second-order effect that investors quietly calculate: each such departure not only removes a brain from Google but also strengthens a competitor. Shazeer takes know-how to OpenAI; Jumper takes it to Anthropic. In a market where a few dozen researchers define the state of the art, losing two to the other side weighs twice as much. It was this reading — talent becoming a rival's advantage — that turned two HR stories into a nearly 7% move in the stock.

The capex bill: why $190 billion spooks the market

Capex (capital expenditure) is the money a company pours into long-term assets — in this case, data centres, TPU chips, and energy to run AI. Alphabet's number for 2026 is not just high: it's vertical.

Year	Capex (approx.)	Note
2022	$31 billion	Pre-generative AI boom level
2025	$91.4 billion	Already doubled in three years
2026 (guidance)	$180–190 billion	About 6x 2022, double 2025
2027	"Significant increase"	No ceiling announced

The company's message is clear: we will spend like never before to avoid falling behind in AI infrastructure. The problem, for those looking at Google shares in the short term, is the timing mismatch. The spending is today, certain and enormous. The return is future, uncertain and hard to measure. When the capex guidance rises and, in the same week, key talent leaves, investors do the less optimistic calculation.

It's worth noting: heavy capex is not necessarily bad news. It's what separates those with the cash to compete from those watching from the sidelines. The negative reading arises when the market doubts the return — and it was precisely that doubt that gained strength in June.

The $84.75 billion equity raise and Buffett's nod

To fund this infrastructure, Alphabet went to the market. And demand was so strong that the offering grew mid-course.

On 1 June 2026, the company announced an $80 billion equity offering. Two days later, on 3 June, it raised the amount to $84.75 billion — because demand exceeded the initial supply. Within the package, one name provides comfort to the narrative: Warren Buffett's Berkshire Hathaway put in $10 billion via a private placement.

There are two possible readings, and the market oscillates between them. The optimistic one: issuing shares with Berkshire on board is a vote of confidence from one of the world's most conservative investors. The pessimistic one: issuing so many shares dilutes existing shareholders, and doing so to fund capex reinforces precisely the fear that the AI bill is too large. Both coexist — and the tension between them is part of what moves Google shares right now.

Google Cloud and Gemini: the side that supports the thesis

It's easy to forget, amid the drop, that the bet has been delivering where it matters most: in the corporate cloud.

In Q1 2026, Google Cloud grew 63% year-on-year, reaching $20.0 billion in revenue, driven by AI solutions and infrastructure for businesses. It's proof that the billion-dollar spend isn't all going down a hole — part is already turning into accelerating revenue. On the consumer side, the Gemini app surpassed 900 million monthly active users, more than double a year earlier.

That's why many analysts treat the drop as a correction, not a break. The infrastructure that scares in capex is the same one delivering 63% Cloud growth. In other words: the two numbers that seem to conflict — record spending and accelerating revenue — are actually two sides of the same strategy. Each new data centre that inflates capex is what allows selling more AI capacity to companies that cannot build their own. The risk is not that the strategy is wrong; it's that the time between paying the bill and reaping the result is too long for market patience. To understand what Google launched and why it fuels the AI thesis, it's worth revisiting what changed with Gemini at Google I/O 2026 and the AI news announced for businesses. The story of Google shares is, at bottom, the story of that bridge between present cost and future revenue.

GOOG vs GOOGL: understanding the ticker before reading headlines

A detail that confuses many people following Google shares: there are two tickers, and they are not the same.

• GOOGL are Class A shares, with voting rights.
• GOOG are Class C shares, without voting rights.

In practice, prices track almost identically, because both represent slices of the same Alphabet. The difference is governance: those buying GOOGL have a voice at meetings; those buying GOOG do not. Headlines often cite one or the other interchangeably — hence the importance of knowing which you are reading. In this post, when we talk about Google shares, the focus is on the stock's behaviour, which moves almost identically in both classes.

What the drop in Google shares teaches about the AI cycle

Even those who don't invest a penny in GOOG have something to learn from this episode. It is a high-resolution snapshot of the current stage of artificial intelligence.

First: AI is genuinely expensive. A capex of $180–190 billion makes explicit that leading this race requires giant cash reserves. For smaller companies, the lesson is not to compete on infrastructure — it's to use well the tools these giants build, without reinventing the wheel.

Second: talent is the bottleneck, not the algorithm. When the departure of two people knocks nearly 7% off the value of one of the world's largest companies, it's clear that applied knowledge is worth more than any isolated model. The same logic applies to any business adopting AI: the advantage lies in those who know how to use it, not just in the technology purchased.

Third: signal and noise get confused in the short term. A one-day drop does not undo 63% Cloud growth. For companies structuring customer service and automation with AI — as happens in any serious AI agent project applied to daily business — what matters is the trajectory, not the week's headline. Google shares will oscillate; business adoption of AI, that only accelerates.

Conclusion: what to watch next

The drop in Google shares in June 2026 is not the end of a thesis — it's the price of ambition being charged in real time. Talent left, the capex bill spooked, and the market asked for a breather. At the same time, Cloud is growing, Gemini is scaling, and Berkshire put money on the table.

What to watch going forward: the next earnings report (and whether capex starts turning into revenue at the promised speed), Google's ability to retain those who stayed, and the pace of Gemini adoption in businesses. If you're evaluating how to bring AI into your business without depending on stock market swings, the way is to start with practical use — and the team at Agathas Web can help design that.

This content is for informational and analytical purposes. It does not constitute a recommendation to buy, sell, or hold any financial asset.