Tech giants pour billions into Anthropic as circular AI investments roll on (arstechnica.com)

Microsoft and Nvidia are doubling down on Anthropic: Nvidia will commit up to $10 billion and Microsoft up to $5 billion as equity investors, while Anthropic pledged a $30 billion commitment to run its models on Microsoft’s cloud as part of a broader partnership. Both companies will also participate in Anthropic’s next funding round. The move binds two major OpenAI backers more tightly to one of OpenAI’s primary rivals; Microsoft says it will both use Anthropic models and provide infrastructure, signaling closer commercial cooperation across competing model makers. Technically and strategically, the deal further concentrates the compute, hardware and cloud relationships that underlie modern LLM development. Anthropic’s massive Azure commitment implies long-term demand for GPU clusters and datacenter capacity for training and inference, while Nvidia’s stake likely ensures preferential access to accelerators and closer co‑development on software/hardware stacks. The announcement follows OpenAI’s own pivot to AWS (a reported $38 billion cloud deal) and its plan for enormous compute scaling (Sam Altman’s stated $1.4 trillion plan for 30 GW), underlining an industry-wide arms race for GPUs, capacity and commercial tie‑ups. That competition could speed model progress but also deepen vendor lock‑in and platform fragmentation, shaping who can compete in frontier AI research and services.