Google is in early talks to deepen its investment in Anthropic

• The talks come as Anthropic and its main rival, OpenAI, are in a multitrillion-dollar race to dominate AI.
• Big Tech companies are taking sides, with Amazon and Google backing Anthropic and Microsoft and Nvidia investing in OpenAI.
• Anthropic and Google declined to comment.

Google is in early discussions to deepen its investment in Anthropic, Business Insider has learned.

The new round of funding could value Anthropic at more than $350 billion, according to one source familiar with the matter. The potential new deal, which is still being negotiated and is in flux, could also take the form of a strategic investment where Google provides additional cloud computing services to Anthropic, a convertible note, or a priced funding round early next year, according to sources familiar with the matter.

It’s unclear how much Google would invest in this round or if there would be additional investors in this latest funding.

Anthropic and Google declined to comment.

Google has invested more than $3 billion in Anthropic, giving it a 14% stake in the company, according to The New York Times, which cited legal filings in March. In September, Anthropic raised $13 billion at a $138 billion valuation, while OpenAI reached a $500 billion valuation last month in a secondary sale that allowed employees to sell shares.

The latest talks come as Anthropic and its main rival, OpenAI, maker of ChatGPT, are in a multitrillion-dollar race to dominate the models that power AI. At the same time, Big Tech companies are taking sides, with Amazon and Google backing Anthropic, while Microsoft and Nvidia have invested billions in OpenAI.

Anthropic was founded in 2021 by former employees of OpenAI, and is best known for its Claude family of large language models.

Google and Anthropic announced a massive cloud computing deal in October worth tens of billions of dollars that gives Anthropic access to up to one million of Google’s custom-designed Tensor Processing Units, or TPUs, which are designed to accelerate machine learning workloads.

Amazon has also invested $14 billion in Anthropic, which uses AWS custom chips to build and deploy its AI models.

Anthropic said last month it remains “committed to our partnership with Amazon, our primary training partner and cloud provider.”

Amazon’s Project Rainer, one of the world’s largest AI compute clusters with nearly half a million Trainium2 chips, recently came online to power Claude.

Anthropic also uses NVIDIA’s GPUs to train and power its models.