Anthropic is expanding its partnership with Amazon through a deal that combines investment with long-term infrastructure use. Amazon is adding another $5 billion, bringing its total investment to $13 billion. In exchange, Anthropic has agreed to spend more than $100 billion on AWS over the next decade, securing the computing power needed to train and run its Claude models at much larger scale.
This follows a similar structure to Amazon’s recent agreement with OpenAI, where funding is tied closely to cloud usage.
A key part of the arrangement is Amazon’s custom AI chips, like Trainium, which are designed to compete with Nvidia. For Anthropic, the priority is reliable access to large-scale compute. For Amazon, the deal ensures sustained demand for its cloud and strengthens its position in the AI infrastructure market.
Why It Matters: What this really shows is just how heavy AI has gotten to run at the highest level. It’s not just about building a better model anymore. You need huge amounts of computing power lined up in advance, and not many companies can actually provide that. It also hints at how tied these AI labs are becoming to specific cloud providers, because once you’re operating at this scale, you don’t have many real alternatives.
- Compute is becoming the limiting factor: It’s no longer enough to have strong models; access to large-scale compute is now a gating issue. Companies that can secure capacity ahead of time can move faster, ship more reliably, and handle spikes in demand, while others may run into delays, outages, or higher costs trying to compete for limited resources.
- Cloud providers are shaping the market: Deals like this tie AI companies closely to specific platforms, which means AWS, Azure, and others are becoming deeply embedded in how AI products are built and delivered. Over time, that influence can extend into tooling, pricing, and even product direction.
- Custom chips are a strategic bet: Amazon is using partnerships like this to push its Trainium chips as a real alternative to Nvidia GPUs. If Anthropic successfully scales on Trainium, it gives Amazon more leverage in the AI stack and could start to shift where developers choose to build.
- Big Tech is hedging its bets: Instead of backing a single winner, companies like Amazon and Microsoft are investing in multiple AI labs. That way, no matter which models or platforms gain the most traction, they still benefit from the underlying infrastructure demand.
- This is about long-term positioning: These agreements lock in relationships, pricing, and capacity for years. That stability can be a major advantage for AI companies trying to grow quickly, while also giving cloud providers predictable, large-scale revenue tied to the future of AI.
Amazon doubles down on Anthropic with $25B investment, mirroring its OpenAI cloud deal – GeekWire
Trusted insights for technology leaders
Our readers are CIOs, CTOs, and senior IT executives who rely on The National CIO Review for smart, curated takes on the trends shaping the enterprise, from GenAI to cybersecurity and beyond.
Subscribe to our 4x a week newsletter to keep up with the insights that matter.
