Google owns a 14 percent stake in Anthropic, court documents seen by The New York Times reveal.

It was previously public knowledge that Google had pumped billions into the generative AI company, but its ownership stake was a secret.

Google has invested more than $3bn into Anthropic across multiple rounds, and is set to invest an additional $750 million in Anthropic in September through convertible debt, according to the filings.

However, the hyperscaler is only allowed to own up to 15 percent of Anthropic. It also has no voting rights, no board seats, and no board observer rights.

The revelation comes after the US Justice Department dropped a proposal to force Google to sell its stakes in generative AI companies, including Anthropic. The company could still be made to report its investments to the DoJ.

Anthropic uses Google Cloud for some of its training and inference work but has more publicly touted its work on Amazon Web Services. Amazon is investing around $8bn into the company (as it came in later, it may not have a larger stake), and is Anthropic's "primary cloud and training partner."

Amazon is also building Rainier, a cluster of Trainium2 UltraServers containing hundreds of thousands of Trainium chips interconnected with third-generation, low-latency petabit scale EFA networking, to be used exclusively by Anthropic.

In its latest funding round earlier this month, Anthropic raised $3.5 billion at a $61.5bn post-money valuation.