Google Bets Up to $40B on Anthropic as AI Compute Race Grows

Google Cloud Capacity Could Help Anthropic Ease Model Growth Constraints

Michael Novinson (MichaelNovinson) •
April 28, 2026    

See Also: AI Impersonation Is the New Arms Race-Is Your Workforce Ready?

The deal expands the Silicon Valley-based search and cloud computing giant’s role as both an investor, infrastructure partner and rival of Anthropic, with Google competing against the San Francisco-based artificial intelligence company through Gemini but also profiting as Anthropic’s Claude platform keeps growing. The Google-Anthropic pact comes days after Amazon committed up to $25 billion to Anthropic.

Google’s investment is being made at a valuation of approximately $350 billion – which aligns with Anthropic’s February funding round – despite secondary market interest suggesting valuations as high as $800 billion, said Bloomberg, which first reported the news. The deal deepens the ties between Google and Anthropic, which includes prior investments exceeding $3 billion and infrastructure collaboration.

A major component of the $40 billion agreement is Google Cloud providing five gigawatts of compute capacity over five years, which highlights how access to infrastructure is central to AI competition. Anthropic relies heavily on Google Cloud and its tensor processing units, which serve as a key alternative to Nvidia chips, giving Google strategic leverage as a supplier of scarce AI compute resources.

Google’s latest investment in Anthropic underscores a broad industry trend where tech titans compete not just on models but on securing compute, chips and energy resources. Google both competes with Anthropic through its Gemini models and supports it as an investor and infrastructure provider, illustrating the complex dynamics in the AI ecosystem. Google confirmed the Bloomberg report to ISMG.

Anthropic has faced growing demand for its services, leading to infrastructure constraints and the need for large-scale compute agreements to sustain growth and performance. Anthropic’s momentum is fueled by strong adoption of its Claude AI models, particularly Claude Code, which has become widely used among developers for software generation. The company’s annual revenue exceeds $30 billion.

Google’s backing on Anthropic is a circular deal in which cloud providers both invest in AI firms and generate revenue by selling them compute resources, according to analysts. The Google deal follows closely behind Amazon’s $5 billion investment in Anthropic – with potential for an additional $20 billion – highlighting intense competition among hyperscalers to back leading AI labs.

Frontier AI development requires tens of billions in capital and long-term infrastructure commitments, which is unchartered territory for the tech industry. The Google-Anthropic deal signals that success in AI is increasingly tied to access to capital, compute and strategic partnerships with major tech firms rather than standalone innovation.

Anthropic is reportedly considering an IPO as early as this October, according to reports, positioning current investments as potentially high-value, pre-public stakes for major backers such as Google. But the company faces potential risks including scrutiny over supply-chain dependencies and concerns about circular financial relationships where investors also serve as infrastructure vendors.

The company also faces a Pentagon designation as a supply-chain risk as well as concerns about misuse of advanced models. The company’s Mythos model disclosed earlier this month highlights Anthropic’s innovation as well as its risk, with restricted access due to cybersecurity concerns and reports of unauthorized use.