Anthropic’s 10GW data center push, Microsoft’s agent defense, and a $6.4B revenue split: latest Anthropic and Microsoft news

Hello,

AI's next phase is being shaped not just by model performance, but by infrastructure decisions, enterprise positioning and the economics flowing through cloud platforms.

Our latest reporting shows how Anthropic and Microsoft are positioning themselves at the center of that shift. Subscribe now for $299 and save 25% on the first year to understand the context behind the headlines, and don't miss the next big story.

The Infrastructure Pivot

Anthropic is assembling former Google data center leaders to secure as much as 10 gigawatts of capacity—an expansion that could cost hundreds of billions of dollars. The move signals a push to reduce reliance on cloud providers and gain more direct control over compute.

The Agent Battle Moves to Work

Anthropic's "Claude Cowork" and other agent tools are designed to operate software autonomously—raising stakes for enterprise incumbents. Microsoft and other software firms are racing to position themselves as the orchestration layer and system of record AI agents must rely on.

Microsoft's Balancing Act

After Anthropic and others pushed deeper into enterprise territory, Microsoft's sales leadership emphasized Azure's security, compliance and multi-model flexibility. Even as AI infrastructure squeezes Azure margins, Microsoft has preserved overall profitability by leaning on high-margin software revenue.

The Economics—and the Risks

Anthropic could share up to $6.4 billion with Amazon, Google and Microsoft in 2027 through revenue-sharing agreements—up sharply from prior years. At the same time, internal research shows Anthropic is studying "rogue agents" and "scheming models" as it prepares to deploy AI into sensitive enterprise environments.

For deeper access to exclusive features like Org Charts for Microsoft and Anthropic, as well as our Deep Research tool and more—subscribe to Pro.