Google Plans Up To $40 Billion Investment In Anthropic, Including $10 Billion Now

Google is preparing to invest up to $40 billion in Anthropic, a move the companies confirmed on Friday as the AI startup accelerates its business behind Claude Code.

“Google is preparing a massive investment in Anthropic, the company behind Claude, despite its own Gemini model competing with the AI company”

9to5Google

TechCrunch reports that the Alphabet subsidiary is committing to invest $10 billion now at a $350 billion valuation for Anthropic, with another $30 billion to follow if Anthropic hits certain performance targets.

9to5Google

CNBC similarly says the pact involves an initial $10 billion from Google at Anthropic's latest valuation of $380 billion, with the remaining $30 billion contingent on certain performance milestones.

The New York Times frames the same structure as $10 billion in cash now and $30 billion later, and says Google will provide five gigawatts of computing power starting in 2027.

Engadget adds that the new investment comes after Anthropic signed a joint agreement with Google and Broadcom for “multiple gigawatts of next-generation TPU capacity.”

Sherwood News ties the expansion to an earlier deal signaled by a Broadcom filing, saying Anthropic would access 3.5 gigawatts of TPU-based AI compute capacity beginning in 2027.

Across the coverage, Anthropic’s access to compute is presented as the core mechanism of the arrangement, with multiple outlets describing Google’s role in supplying TPUs and cloud capacity.

The investment is portrayed as part of a broader scramble for the compute needed to train and deploy frontier AI systems, with TechCrunch saying “The AI race is increasingly defined by access to the compute needed to train and deploy these systems.”

TechCrunch reports that Anthropic has been in a scramble of its own and points to a deal with CoreWeave for data center capacity earlier this month, while also noting that this week it secured an additional $5 billion investment from Amazon.

About Amazon

CNBC similarly describes how Anthropic’s latest deal with Google lands as the company works to address “inevitable strain” on its infrastructure from growing enterprise, developer and consumer demand for Claude.

The New York Times adds that Anthropic relies on Google and Amazon for the cloud computing it needs to build and run its Claude chatbot and coding systems, and that it is also one of the largest users of custom AI chips including Google’s tensor processing units, or TPUs, and Amazon’s Trainium.

Sherwood News says the Google expansion is an expansion of the deal between Google, Broadcom, and the Claude developer revealed earlier this month, and it specifies that Broadcom signaled Anthropic would access 3.5 gigawatts of TPU-based AI compute capacity beginning in 2027.

Sherwood News then states that “Google Cloud will now be providing 5 gigawatts over the next five years,” and it adds that the company could provide more thereafter.

Engadget emphasizes the structure by saying Google will also provide 5 gigawatts of computing capacity in 2027 through Anthropic’s existing commitment to use Google’s TPUs and servers.

Several outlets describe the Google-Anthropic arrangement as a “circular” pattern in which investment and infrastructure flow together, and Engadget explicitly frames it that way by saying the investment could be viewed as a circular deal with the AI startup (and frequent competitor).

Engadget notes that Google has invested in Anthropic at multiple points in the past, and it ties the new investment to Anthropic’s joint agreement with Google and Broadcom for “multiple gigawatts of next-generation TPU capacity.”

TechCrunch adds that Anthropic’s relationship with Google predates this week’s news, describing an earlier partnership announced earlier this month with Google and chipmaker Broadcom to access multiple gigawatts of TPU-based computing capacity beginning in 2027, and it reports that a subsequent Broadcom securities filing put that figure at 3.5 gigawatts.

CNBC provides additional historical detail, saying the relationship dates back to 2023, when Google invested $300 million in the AI lab for a stake of about 10%, and that months later Google poured in another $2 billion.

CNBC also says ahead of Friday’s announcement, Google’s investment in Anthropic exceeded $3 billion and it reportedly owned a 14% stake in the company.

The New York Times describes the broader ecosystem by saying Google and Amazon began backing Anthropic in its infancy and have become two of its largest minority investors, while also noting that the agreements are examples of how circular deals have become critical to the AI boom.

In parallel, TechCrunch and Engadget both point to Amazon’s earlier commitments, with TechCrunch saying Amazon secured an additional $5 billion investment from Amazon and Engadget recalling that earlier in April Amazon announced it would invest $5 billion in Anthropic with an additional $20 billion in payments available if certain milestones were met.

While the investment announcement itself is largely described through corporate and financial reporting, CNBC includes a direct statement from Anthropic’s CEO Dario Amodei about why the infrastructure matters, quoting, “Our users tell us Claude is increasingly essential to how they work, and we need to build the infrastructure to keep pace with rapidly growing demand.”

TechCrunch describes Anthropic’s approach to access control for its models, saying “Due to potential misuse, Anthropic has restricted broader access while it works with select organizations to evaluate and address those risks,” and it adds that the model has “already fallen into unsanctioned hands.”

Bitcoin News

The New York Times connects the deal to product momentum, saying it comes as the AI start-up looks to keep up with accelerating demand for its business and coding products, and it specifies that the deal is tied to Claude Code.

CNBC also situates the announcement in the competitive landscape by noting that Google provides access to Anthropic's Claude models through its cloud division and that Google’s Gemini is competing with Anthropic in the market for AI models and services.

TechCrunch says Google is a direct competitor in AI models but also a key infrastructure supplier to Anthropic, and it describes Anthropic’s reliance on Google Cloud for TPUs.

The New York Times adds a financial and adoption dimension, stating that this month Anthropic said its annual revenue run rate surpassed $30 billion, up from $9 billion at the end of 2025.

It also notes that Google has its own competing AI products, with Gemini powering Google Search and a separate chatbot and research system that subscribers pay to use.

The outlets differ in emphasis even when describing the same core commitment, and those differences show up in how they frame valuation, compute timing, and the broader significance of the deal.

TechCrunch focuses on the AI race for compute and says Anthropic released its latest model, Mythos, to a limited group of partners this month, adding that Mythos is “the company’s most powerful model to date” and has “significant cybersecurity applications.”

bloomingbit

CNBC instead emphasizes strategic investment and provides a valuation of $380 billion for Anthropic at the time of the pact, while also detailing Google’s earlier stakes and ownership, including that it reportedly owned a 14% stake ahead of Friday’s announcement.

The New York Times highlights the business and coding demand behind Claude Code and includes a specific comparison about electricity, stating that the five gigawatts of computing power starting in 2027 is “enough electricity to power all the homes in Minnesota.”

Engadget foregrounds the “circular deal” framing and points to the similarity with Amazon’s earlier investment structure, while 9to5Google stresses that the investment will support a “significant expansion” to Anthropic’s computing capacity and reiterates that Google is one of the few companies with access to Anthropic’s Mythos model.

Sherwood News, meanwhile, concentrates on the expansion mechanics by citing Broadcom’s earlier signaling of 3.5 gigawatts and then describing Google Cloud providing 5 gigawatts over the next five years.

Even when they agree on the headline number of “up to $40 billion,” the reporting diverges on the valuation figure and on how much of the story is about product demand versus infrastructure capacity.

The reporting suggests several near-term implications for Anthropic and for the companies supplying its compute, with the deal structured around milestones and capacity delivery.

TechCrunch says the promise of investment comes after Anthropic released Mythos to a limited group of partners this month, and it adds that Anthropic has restricted broader access due to “potential misuse,” even as the model has “already fallen into unsanctioned hands.”

It also notes that Anthropic’s valuation stood at $350 billion as recently as February and says investors have since been eager to back the company at $800 billion or more, according to Bloomberg, while also saying the company is reportedly considering an IPO as soon as October.

CNBC describes the operational pressure behind the investment by saying Anthropic is working to address “inevitable strain” on its infrastructure from growing demand for Claude, and it ties Google’s cloud and TPUs to keeping pace with that demand.

The New York Times adds that the computing commitment could help Anthropic as it tries to build out the infrastructure it needs to meet the surging demand for its Claude Code, and it states that companies are signing up to use Anthropic’s coding and business tools to develop programs and applications.

Engadget and TechCrunch both emphasize that the deal expands an existing arrangement with Google and Broadcom for TPU capacity beginning in 2027, with Engadget saying Google will provide 5 gigawatts of computing capacity in 2027 and TechCrunch describing Google Cloud providing a fresh 5 gigawatts of capacity over the next five years.

Sherwood News similarly says Google Cloud will now be providing 5 gigawatts over the next five years, and it adds that it could provide more computing power if needed.