Google revealed to hold 14% stake in Anthropic

Tech giant is set to invest a further $750m in September

Google has contributed $3 billion to Anthropic over multiple funding rounds

Google owns a 14% stake in Anthropic, according to court documents obtained by The New York Times, shedding new light on its financial involvement with the frontier AI firm.

While it was widely known that Google had invested billions in Anthropic, as previously reported by Computing, the precise extent of its ownership had remained undisclosed. The filings reveal that the tech giant has contributed over $3 billion across multiple funding rounds and is set to invest a further $750 million in September via convertible debt.

Despite this substantial financial backing, Google is restricted to a maximum 15% stake in the company. It holds no voting rights, board seats, or board observer privileges.

The disclosure comes in the wake of the US Department of Justice (DoJ) proposing that Google divest its stakes in generative AI firms, including Anthropic. However, the company may still be required to report such investments to regulators.

Claude family of AI models

Anthropic, which develops the Claude family of AI models, utilises Google Cloud for certain training and inference workloads but has publicly aligned itself more closely with Amazon Web Services. Amazon has pledged up to $8 billion in investment and serves as Anthropic’s primary cloud and training partner.

As part of this partnership, Amazon is developing Rainier, a high-performance AI cluster comprising hundreds of thousands of Trainium2 UltraServer chips, interconnected via third-generation petabit-scale EFA networking, exclusively for Anthropic’s use.

Earlier this month, Anthropic secured $3.5 billion in fresh funding, valuing the company at $61.5 billion post-investment.