Google will continue to offer Anthropic's artificial intelligence models to customers via Google Cloud, with exceptions for defense-related projects. This decision follows the US Department of Defense designating Anthropic as a supply chain risk.
Anthropic's Claude models are accessible through Google Cloud's Vertex AI platform. Google, a significant investor in Anthropic, recently committed an additional $1 billion in January 2025, building on a prior $2 billion stake. Anthropic utilizes Google Cloud infrastructure for model training and has expanded its partnership, gaining access to one million of Google's custom tensor processing units.
The Pentagon's designation stemmed from Anthropic's refusal to accept new terms for its AI systems from the US Department of Defense. Previously, President Donald Trump directed federal agencies to halt Anthropic technology use, with Defense Secretary Pete Hegseth stating a six-month phase-out of Pentagon work with the company.
Microsoft has also confirmed it will continue supporting Anthropic products for non-Department of War clients after its legal review. Anthropic CEO Dario Amodei intends to contest the government's designation in court. Amazon also confirmed it will continue offering Anthropic's AI technology to its cloud customers, excluding Department of Defense work.