Alphabet’s $80 Billion Capital Raise Backs a $35 Billion Anthropic Wager Amid Legal and Operational Challenges

Alphabet orchestrated one of the most consequential days in its corporate history on June 10, unveiling an $80 billion capital increase — the largest ever for the company — while simultaneously anchoring a $35 billion chip-and-infrastructure pact with AI startup Anthropic. The twin moves, together with a freshly raised capex target of $180 billion to $190 billion for 2026, lay bare the scale of the Google parent’s ambition to dominate the AI race. But the day also brought headwinds: a stinging legal setback and a data center fire that disrupted cloud services in India.

The capital raise itself is a multi-layered financial operation. Berkshire Hathaway, Warren Buffett’s investment powerhouse, agreed to purchase $10 billion in Alphabet shares through a private placement — a rare public vote of confidence from the famously tech-averse conglomerate. A further $30 billion will be raised through public offerings split between preferred and common stock. The remaining $40 billion is set to flow through an at-the-market (ATM) program scheduled to launch in the second half of 2026. The infrastructure splurge is the clear priority: roughly $50 billion of the proceeds are earmarked for AI data centers, chips, and global computing capacity, while $30 billion will cover tax liabilities tied to employee stock compensation.

Alongside the capital raise, Alphabet deepened its ties to Anthropic by underwriting a 35-billion-dollar chip arrangement. The deal goes far beyond a simple financing: Google is guaranteeing lease payments for five Anthropic-operated data centers and supplying its own Tensor Processing Units to the startup. The effect is to bind Anthropic’s platform ever more tightly to Google Cloud, ensuring that developers using Anthropic’s models will increasingly run on Google hardware. The partnership cements Alphabet’s role as both financier and infrastructure backbone for one of the most prominent firms in the generative AI race.

The legal troubles emerged from a California courtroom, where a judge rejected Alphabet and Meta’s request for a new trial in a social-media addiction lawsuit. A jury had previously found YouTube 30% liable — corresponding to $1.8 million of the total $6 million in damages awarded — while Meta carried the remaining 70%. Both companies plan to appeal, but the ruling ensures the litigation will remain a persistent distraction.

Should investors sell immediately? Or is it worth buying Alphabet?

Operational disruptions added to the day’s complexity. A fire broke out at a Google data center in Delhi, triggering an emergency shutdown of networking infrastructure. Users in Delhi, Chennai, and Mumbai reported elevated latency on Google Cloud services, and engineers were dispatched to restore normal operations.

On the product front, Alphabet’s in-house AI efforts continued to gain traction. The Gemini app now boasts 900 million monthly active users — double the previous count — and the company is rolling out new generative features tailored to small businesses. DeepMind also released DiffusionGemma, a model built on a diffusion-based architecture that generates text in blocks rather than sequentially, capable of more than 1,000 tokens per second on Nvidia H100 hardware. The model is especially well-suited for code completion, though output quality trails the established Gemma 4.

Alphabet shares closed at €309.15 on Wednesday, roughly 12% below the all-time high of €350.75 reached in mid-May. On a 12-month view, the stock has more than doubled, gaining about 101%. The relative strength index hovered around 41, signaling a neutral-to-slightly-weak market posture — no buying frenzy, but no panic either. The fundamental AI narrative remains intact, even as legal clouds and operational hiccups test near-term sentiment.