Amazon has launched a new offering called “AI Factories,” enabling large enterprises and government agencies to run Amazon’s AI systems within their own data centers. Under this model, customers provide the physical infrastructure and power, while AWS delivers and manages the AI system, seamlessly integrating it with other AWS cloud services like Amazon Bedrock, SageMaker, and core networking, storage, and security tools. The initiative targets organizations with strict data sovereignty requirements—those needing full control over their data to prevent it from being accessed by competitors or foreign governments. By deploying AI Factories on-premises, companies avoid sending sensitive data to external cloud providers or sharing hardware with third parties. The name “AI Factory” is not coincidental. It mirrors Nvidia’s own branding for its high-performance AI hardware systems, which bundle GPUs, networking, and software tools into turnkey solutions for AI workloads. AWS’s AI Factory is a joint effort with Nvidia, combining AWS’s cloud infrastructure with Nvidia’s latest hardware, including its Blackwell GPUs. Customers also have the option to use Amazon’s Trainium3 chips, the company’s custom AI accelerators. This move marks a strategic shift for AWS, which has long championed public cloud adoption. Now, in response to growing demand for data control and regulatory compliance, AWS is embracing hybrid and private deployments. The solution supports local data processing, enhances security, and allows organizations to maintain compliance with national data laws. Amazon is not alone in this trend. Microsoft has also rolled out AI Factories across its global network to power OpenAI workloads, building what it calls “AI Superfactories” in locations like Wisconsin and Georgia. These facilities rely heavily on Nvidia’s AI infrastructure. While Microsoft initially focused on its public cloud AI deployments, it has since expanded options for private environments through Azure Local—a managed hardware solution that can be installed on customer premises. The resurgence of on-premises AI infrastructure echoes a pattern from the early 2010s, when cloud adoption was still emerging. Now, as AI becomes central to business and national strategy, even the largest cloud providers are investing heavily in private data centers and hybrid models. This shift underscores a growing tension: while cloud computing promised scalability and efficiency, the rise of AI has reignited demand for control, security, and localized processing—bringing the industry full circle.