Recursive Superintelligence Commits $410M to AWS

Recursive Superintelligence (Recursive), the AI startup founded by former Salesforce AI chief Richard Socher, has signed a $410 million multi-year compute agreement with Amazon Web Services (AWS), marking one of the largest cloud infrastructure commitments by a young AI company this year. The deal comes just two months after Recursive emerged from stealth with a $650 million funding round, underscoring how quickly AI startups are converting fresh capital into computing power.

Compute is becoming the industry's biggest competitive advantage

Rather than building its own data centers, Recursive will use AWS as its primary cloud provider to train and run its self-improving AI research system. The company says its platform is designed to automate scientific discovery by continuously improving its own research process through recursive learning, a concept many AI labs see as a potential path toward more capable AI systems.

The agreement highlights a broader shift in the AI industry, where access to large-scale compute is increasingly becoming as valuable as talent or funding. As frontier models grow larger, infrastructure spending is emerging as one of the biggest expenses for AI developers.

AWS continues adding high-profile AI customers

Recursive joins a growing list of AI companies expanding their infrastructure on AWS. Earlier this year, Anthropic broadened its long-term partnership with Amazon, securing up to 5 gigawatts of compute capacity, while AWS also activated Project Rainier, one of the world's largest AI compute clusters built around nearly 500,000 Trainium2 chips.

The latest agreement further strengthens AWS's position in the race to supply computing infrastructure for frontier AI labs, where cloud providers are increasingly competing on specialized AI hardware and large-scale capacity rather than traditional cloud services alone.

AI investment is increasingly becoming compute investment

The announcement also reflects a growing trend across the AI industry: raising capital is only the first step, while securing long-term compute has become equally important.

Over the past year, several leading AI companies have announced multibillion-dollar infrastructure partnerships as demand for GPUs and AI accelerators continues to outpace supply. Instead of purchasing hardware outright, startups are increasingly signing long-term cloud agreements that guarantee access to scarce computing resources.

Recursive's decision to dedicate more than half of its recently raised capital toward AWS infrastructure signals how essential compute has become for companies developing next-generation AI systems. As competition shifts from building models to scaling them, long-term access to computing capacity is rapidly becoming one of the industry's most strategic assets.