Amazon AI Capacity Shortage Drives Customers to Google, Other Competitors
November 24, 2025 – A critical shortage of artificial intelligence computing capacity at Amazon Web Services (AWS) has prompted numerous customers to shift their AI workloads to rival cloud providers, moast notably Google Cloud, according to reports surfacing today.The disruption, impacting businesses developing and deploying large language models and other AI applications, underscores the escalating demand for AI infrastructure and the challenges cloud providers face in keeping pace.
The capacity constraints at AWS, first reported by The Information, stem from unexpectedly high demand for AI chips, notably GPUs, and a slower-than-anticipated expansion of Amazon’s AI infrastructure.This situation has created a bottleneck for companies reliant on AWS for their AI initiatives,forcing them to explore alternatives to avoid delays in product launches and research. The shift highlights the growing competitive landscape in the cloud AI market, where Google and microsoft are aggressively investing to capture market share.
Several companies, including Stability AI, have publicly acknowledged difficulties securing sufficient GPU capacity from AWS. Stability AI CEO Emad Mostaque told The Information that his company was “actively diversifying” away from AWS due to the limitations, stating the situation was “very serious.” Other affected customers, speaking on background, reported experiencing significant wait times for GPU instances and, in some cases, being unable to obtain the resources needed for their projects.
The issue centers around the availability of high-end GPUs, such as nvidia’s H100 chips, which are essential for training and running large AI models. While AWS has been working to increase its supply of these chips, demand has outstripped availability, leading to the current shortage. Google Cloud has been actively marketing its own AI infrastructure, including its Tensor Processing Units (TPUs), as a viable alternative for customers facing challenges with AWS.
Amazon acknowledged the capacity constraints in a statement, saying it is “working hard” to increase availability and is “investing heavily” in AI infrastructure. The company stated it is indeed prioritizing existing customers and working to onboard new ones as quickly as possible.However, the situation is expected to persist for the near term, potentially giving Google and other competitors a significant advantage in attracting AI-focused businesses.
