Google invests up to $2 billion in Anthropic

Google agreed to invest up to $2 billion in Anthropic, TechCrunch and other outlets reported, with $500 million paid immediately and a further $1.5 billion to follow over time. The deal was structured as a convertible note, debt that converts to equity at Anthropic’s next priced funding round, and neither company issued its own press release confirming the terms; reporting traced to people familiar with the deal, first via the Wall Street Journal.

It was Google’s second disclosed investment in Anthropic that year. The company had already put $300 million into Anthropic that February for roughly a 10% stake, alongside a cloud-computing partnership under which Anthropic would train and run its models on Google Cloud’s GPU and TPU clusters. Google was also reported to have participated, alongside Salesforce and others, in Anthropic’s $450 million Series C that May.

The commitment came almost exactly a month after Amazon’s own pledge of up to $4 billion to Anthropic, which made AWS Anthropic’s primary cloud provider and committed the lab to Amazon’s Trainium and Inferentia chips. Coverage at the time framed the two deals as a “proxy war” between the cloud providers, each securing a stake in a leading AI lab and a flagship customer for its own infrastructure without fielding a comparably positioned frontier model of its own — Google’s Gemini family had not yet launched. The arrangement gave Anthropic two major hyperscaler backers simultaneously, a structure that recurred and deepened across further rounds through 2024 and 2025.