Google to Invest up to $40 Billion in AI Company Anthropic
- The setup: an AI lab everyone wants a piece of
- Amazon moves first
- Google answers with a bigger check
- Cash, chips, and an invisible moat
- The Mythos catalyst
- A market rewired around compute
- Google’s long AI play pays off
- Frenemies, everywhere
- What changes next
Google to Invest up to $40 Billion in AI Company Anthropic Human Human coverage presents Google’s plan to invest up to $40 billion in Anthropic as a staged deal anchored by an initial $10 billion, blending cash with large allocations of Google Cloud compute at a roughly $350 billion valuation. It highlights how this move, alongside Amazon’s multibillion-dollar commitment, underscores an AI arms race among tech giants while deepening Anthropic’s reliance on Big Tech infrastructure and inviting potential regulatory scrutiny. @Arstechnica @Verge @TC @4qd8…qnwa Google has turned the AI arms race into an all‑out compute land grab, betting up to $40 billion on Anthropic—and in the process blurring the line between rival and customer in the most expensive corner of tech.
The setup: an AI lab everyone wants a piece of
Anthropic, the safety‑branded AI lab behind the Claude models, has spent the past year sprinting to keep up with demand. Its Claude assistants, Claude Code for developers, and newer tools like Claude Cowork have seen “rapid growth,” pushing the company into outages, strict usage limits, and experiments with throttling heavy features just to stay online.1
At the same time, OpenAI has been loudly telling investors that the future belongs to whoever controls the most compute. Co‑founder Greg Brockman framed it bluntly: “We are moving to a compute-powered economy,” boasting OpenAI expects 30 gigawatts of compute by 2030—far ahead of what Anthropic can access today.2
Into that gap stepped Big Tech.
Amazon moves first
In the days before Google’s blockbuster announcement, Amazon quietly escalated its own bet on Anthropic. It had already put $8 billion into the lab, then added another $5 billion with a path to invest “up to an additional $20 billion in the future.”3 The deal goes far beyond equity: Anthropic agreed to spend $100 billion over time to lock in around 5 gigawatts of compute capacity from Amazon to train and run Claude.2
That arrangement, as Axios drily notes, is part of an “AI financing circle” in which cloud giants bankroll AI labs, which then turn around and commit colossal sums right back to those same clouds for capacity.2
Anthropic’s compute scramble didn’t stop at Amazon. Earlier in the month, it also struck a deal with CoreWeave for additional data center capacity, another attempt to tame the usage spikes that were forcing Claude users into limits and outages.4
Google answers with a bigger check
On April 24, the other shoe dropped. Bloomberg first reported—then outlets from The Verge to TechCrunch and Axios confirmed—that Google would pump at least $10 billion into Anthropic, with the potential to go all the way to $40 billion if performance targets are hit.3421
The initial $10 billion values Anthropic at roughly $350 billion, putting a still‑private AI lab in the same neighborhood as established megacaps.4 The remaining $30 billion is contingent—essentially a huge performance‑based tab Google can pick up if Anthropic keeps executing.3
As The Verge sums it up, “Google is investing billions in Anthropic.”3 But the headline number is only half the story.
Cash, chips, and an invisible moat
TechCrunch describes the deal as “up to $40B in Anthropic in cash and compute,” underscoring what actually matters: Anthropic isn’t just getting money; it’s getting a firehose of Google Cloud capacity.4
Anthropic “relies heavily on Google Cloud for chips and infrastructure, including access to Google’s tensor processing units (or TPUs), which are specialized chips designed for AI workloads and considered among the best alternatives to Nvidia’s in-demand processors.”4 As part of the new deal, Axios reports, Anthropic will also get access to 5 gigawatts of compute from Google—matching its long‑term commitment with Amazon.2
This isn’t their first dance. Earlier in the month, Anthropic announced a partnership with Google and chip designer Broadcom to secure additional specialized hardware—part of a broader pattern in which “established companies like Microsoft have products and services that can help new AI companies like Anthropic scale, so the former invests in the latter so the latter can, in turn, pay for the former’s products and services.”4
Ars Technica notes that this structure—equity plus long‑term cloud lock‑in—has effectively become the “common scheme for investment in AI companies like Anthropic.”4
The Mythos catalyst
Timing matters. Google’s investment promise comes on the heels of Anthropic’s release of Mythos, billed as its “most powerful model to date” with “significant cybersecurity applications.”4
Anthropic has restricted Mythos to a limited group of partners “due to potential misuse,” and is working with select organizations to evaluate risks—even as the model has already slipped into “unsanctioned hands,” according to TechCrunch.4 It’s also “likely expensive to run at scale,” driving home why Anthropic is racing to bolt on so much new capacity.4
Behind Mythos is a broader shift in how people use AI: more “agentic workflows,” more automation, and more specialized tools like Claude Code and Claude Cowork. Ars Technica points out that demand ranges from “big improvements to setbacks” depending on how these tools are integrated, but the net effect has been a “dramatic increase in demand for Anthropic’s services.”4
A market rewired around compute
The result is that the AI race is no longer just about who has the best model; it’s about who can afford to feed it. TechCrunch notes that “the AI race is increasingly defined by access to the compute needed to train and deploy these systems,” with OpenAI “moving aggressively to secure that capacity through a web of multi-hundred-billion deals across cloud providers, chip suppliers, and energy.”4
Greg Brockman’s “compute-powered economy” remark frames this bluntly: whoever controls the most energy‑hungry data centers wins.2
Google and Amazon’s dueling Anthropic checks are the visible tip of that shift. Axios puts it simply: “AI labs will keep making deals to get compute, and those providing it will keep betting on different labs to succeed until the race is over.”2
Google’s long AI play pays off
Inside Google, the deal looks less like a pivot and more like the logical next chapter in a decade‑long bet.
Demis Hassabis, who runs Google DeepMind, amplified a reminder that back in 2016, CEO Sundar Pichai declared Google would be an “AI‑first company,” backing “a series of projects—custom chips, Cloud, YouTube, and deep AI research—that seemed to have nothing to do with Google’s core search product. All of these bets have paid off.”5
Pichai himself has been crowing about the payoff. In an April earnings‑season post, he wrote that 2026 is “off to a terrific start,” saying Google’s “AI investments and full stack approach are lighting up every part of the business,” with search at an all‑time high, Google Cloud revenue up 63 percent, and Gemini models showing “incredible mo…”—a clear victory lap tying infrastructure, models, and now external customers like Anthropic into one narrative.6
The Anthropic deal further validates that strategy: TPUs, Cloud, and AI research are no longer side bets—they’re the leverage Google uses to pull an independent frontier lab into its orbit, even while competing head‑on in products.
Frenemies, everywhere
There’s an almost surreal tension at the heart of all this. Google and Amazon both build their own frontier models and AI assistants. Yet both are “big backers of Anthropic, despite their own AI efforts.”2
Anthropic, meanwhile, is simultaneously a customer, a portfolio company, and a rival to all of them.
Axios calls Google’s move “Big Tech’s latest huge AI bet,” emphasizing how these giants now treat independent labs both as competitive threats and indispensable partners.2 Ars Technica notes that “this is not the first time Anthropic has received investment from Google, even though Google is ostensibly competing with Anthropic over AI models.”4
The Verge cuts to the chase: between Amazon’s and Google’s layered deals, “tech companies with their own AI ambitions keep betting on Anthropic.”3
What changes next
In the near term, Anthropic’s customers may simply see fewer outages, looser limits, and more powerful models like Mythos and Claude Code reachable through APIs instead of invite‑only programs.4 The fresh cash and contracted compute are explicitly meant to “help close the gap between demand and supply” for Anthropic’s tools.4
In the medium term, though, the market is consolidating into a tight web of AI labs and cloud oligopolies, each bound together by decade‑long commitments and multi‑billion‑dollar credit lines. Deals like Google’s $40 billion Anthropic play don’t just finance today’s training runs; they pre‑allocate tomorrow’s energy, chips, and data center capacity to a small club.
Everyone else—startups, open‑source projects, even national AI programs—will be competing in the shadow of those gigawatts.
1. Google will invest as much as $40 billion in Anthropic — Ars Technica on Anthropic’s rapid growth, outages, and scaling challenges.
2. Google’s $40B Anthropic move is Big Tech’s latest huge AI bet — Axios on Google’s valuation, gigawatt commitments, and Greg Brockman’s “compute-powered economy” line.
3. Google is investing billions in Anthropic. — The Verge summarizing Google’s initial $10B and potential $30B follow‑on, plus Amazon’s parallel investment.
4. Google to invest up to $40B in Anthropic in cash and compute — TechCrunch on the $40B structure, Mythos, compute competition, and CoreWeave and Amazon deals.
5. @demishassabis on X — Retweet highlighting Pichai’s 2016 “AI-first company” strategy and how those bets have “paid off.”
6. @sundarpichai on X — Pichai touting a “terrific” 2026 start, with AI investments and a full‑stack approach boosting Search and 63% Cloud revenue growth.
Story coverage
Write a comment