Amazon bets on AI: Andy Jassy sees AWS reaching $600 billion in annual revenue

Amazon CEO Andy Jassy told employees that in ten years the AWS cloud division could reach $600 billion in annual revenue. The core of the forecast is artificial intelligence, which the company sees as the next major accelerator of its cloud business.

The AWS forecast

The statement came at an internal meeting with employees and looks like one of the most ambitious targets publicly tied to the cloud market. Jassy is effectively talking not just about AWS growing, but about moving to a different scale, where AI is not an extra service but the central reason companies increase infrastructure spending. For Amazon this is an important signal both inside the company and for investors: management believes demand for compute, data storage and AI services is only beginning to take off.

The context is telling too. The market already hears enormous estimates of future AI revenue, and such forecasts are gradually becoming part of the competitive race among the largest tech companies. Where the cloud was once sold as a convenient way to rent servers and databases, it is now increasingly pitched as the foundation for training, deploying and running AI models. By this logic, AWS should earn not only from classic infrastructure but at every stage of the AI chain.

Why bet on AI

Amazon's logic is clear: a business that adopts generative models and AI agents almost always consumes more cloud resources. It needs accelerators, network capacity, storage, data preparation tools, security services and platforms for running models in production. The more companies move their AI experiments into real processes, the better the odds that cloud providers can turn one-off interest into recurring revenue. For AWS, this is a chance to sell not a single product but an entire stack. Amazon can earn on this bet at several layers at once:

• rental of compute for training and inference of models
• storage and processing of corporate data for AI scenarios
• tools for deploying, monitoring and scaling models
• services for security, access control and compliance with internal policies
• integration of AI features into existing AWS cloud products

That is why the $600 billion forecast looks less like a bet on one trendy service and more like a calculation based on a systemic shift in corporate IT budgets. If companies build internal assistants, analytics tools, search systems and customer-service automation on top of the cloud, the provider collects money at several points, from hardware to application services. This model especially favors players that already have a huge base of corporate customers.

How big is the scale

The $600 billion figure matters not only as a headline-grabbing marker. It shows how large players estimate the potential size of AI infrastructure in ten years. Essentially, Amazon is working from a scenario in which artificial intelligence becomes as basic a line item for business as databases, compute clusters, analytics and network services are today. This is no longer talk of experimental pilots but of a long investment cycle that assumes mass adoption.

But so large a forecast automatically raises the bar for execution. To come close to such numbers, AWS will have to hold its cloud leadership, expand its data centers, secure access to expensive chips and at the same time withstand pressure from other providers that are also building AI platforms. In addition, corporate customers will expect not only capacity but clear project economics: if adopting AI does not cut costs or add revenue, they will not keep growing their cloud bills indefinitely.

What it means

Jassy's forecast shows that the largest cloud platforms no longer treat AI as a separate vertical. For Amazon it is the future engine of all of AWS, and for the market it is a sign that the contest is no longer just over models but over control of the entire infrastructure those models run on.