Google invests up to $40 billion in cash and compute in Anthropic

Google plans to invest up to $40 billion in Anthropic to support the AI ​​company’s growing computing needs, Bloomberg reports. According to Anthropic, the Alphabet subsidiary currently plans to invest $10 billion at Anthropic’s valuation of $350 billion, and an additional $30 billion if Anthropic meets certain performance goals.

The investment commitment comes after Anthropic released its latest model, the Mythos, to a limited group of partners this month. Anthropic says Mythos is its most powerful model to date and has important cybersecurity applications. Due to the potential for abuse, Anthropic has restricted widespread access and is working with some organizations to assess and address the risks, but this model is already in unauthorized hands. It can also be expensive to run at scale.

The AI ​​race is increasingly defined by access to the computing needed to train and deploy these systems. OpenAI is moving aggressively to secure its capabilities through hundreds of billions of dollars in deals across cloud providers, chip suppliers and energy, including this month’s expanded deal with chipmaker Cerebras.

Anthropic is in its own mess. The company has faced widespread complaints about Claude’s usage restrictions in recent weeks, and responded with a series of infrastructure deals. Earlier this month, Anthropic signed a deal with cloud computing provider CoreWeave for data center capacity. Also this week, the company secured another $5 billion in investment from Amazon. The deal is part of a broader agreement under which Anthropic will spend up to $100 billion on about 5 gigawatts of computing power over the long term.

Although Google is a direct competitor in AI models, it is also an important infrastructure supplier to Anthropic. Anthropic relies heavily on Google Cloud for chips and infrastructure, including access to Google’s tensor processing units (TPUs). TPUs are specialized chips designed for AI workloads and are considered one of the best replacements for Nvidia’s high-demand processors.

Anthropic’s relationship with Google predates this week’s news. Earlier this month, Anthropic announced a partnership with Google and chipmaker Broadcom, which designs custom AI chips for Google, giving it access to gigawatts of TPU-based computing capacity starting in 2027. Subsequent Broadcom securities filings put that figure at 3.5 gigawatts.

Google’s new investment expands that deal, giving Google Cloud an additional 5 gigawatts of capacity over the next five years and room for further expansion.

Anthropic was valued at $350 billion in February. Investors have since been eager to back the company with more than $800 billion, according to Bloomberg. The company is reportedly considering an IPO as early as October.