Anthropic partners with Google Cloud to scale its AI systems

Anthropic announced a partnership with Google Cloud on February 3 to support the training and deployment of its AI systems. The companies plan to co-develop AI computing systems, with Anthropic using Google’s graphics processing unit (GPU) and tensor processing unit (TPU) clusters.

Photo: Anthony Quintano, “Google at the 2012 Democratic National Convention North Carolina”. Taken 6 September 2012. CC BY 2.0. Resized and cropped. Historical Google display; no endorsement implied.

Google Cloud says Anthropic has selected it as its preferred cloud provider. It plans to build large-scale, next-generation GPU and TPU clusters for Anthropic’s training and deployment work.

Anthropic CEO Dario Amodei said the partnership would give the company the computing performance and scale needed to bring its AI systems to more people. Anthropic’s developing Claude, an assistant based on its research into safer, more steerable AI.

That work includes Constitutional AI, a safety technique Anthropic says can help make systems easier to understand and rely on.

Google says Claude already runs on Google Cloud and that Anthropic is working with early partners, with broader access planned in the coming months. This partnership announcement doesn’t open unrestricted public access to Claude.