Amazon: AI is big business

Amazon just put a number on a business most big tech companies still describe in foggy language: its Amazon Web Services artificial intelligence services are running at more than $15 billion a year in the first quarter of 2026. That was the first time Amazon gave investors a direct revenue figure for that effort. (reuters.com, aboutamazon.com) That number matters because Amazon Web Services is not a side project inside Amazon. It is the cloud division that rents computing power and software to other companies, and it has become one of Amazon’s main profit engines over the past decade. (aboutamazon.com, cnbc.com) Andy Jassy paired the $15 billion figure with another one: Amazon’s chips business, which includes Graviton and Trainium processors, is now running at more than $20 billion a year. Those chips are the hardware Amazon uses to lower the cost of running artificial intelligence workloads inside its own cloud. (reuters.com, aboutamazon.com) The simple version is that Amazon is trying to own more of the stack. Instead of paying outside suppliers for every expensive artificial intelligence chip, it wants to design more of the silicon itself and then rent the whole system to customers through Amazon Web Services. (aboutamazon.com, reuters.com) Jassy used his annual shareholder letter on April 9, 2026 to argue that this is not spending on faith alone. He wrote that much of Amazon’s artificial intelligence infrastructure investment already has customer demand behind it, which is his way of telling investors these data centers are being built for orders that already exist. (aboutamazon.com, reuters.com) That helps explain the talk around a roughly $200 billion investment push. Reports tying together Jassy’s comments and Amazon’s plans describe a multi-year buildout aimed at more data centers, more custom chips, and more artificial intelligence tools across both cloud computing and Amazon’s shopping business. (infotechlead.com, cnbc.com) Amazon is also selling artificial intelligence in layers. At the bottom it rents raw computing power, in the middle it offers models and developer tools through services like Bedrock, and at the top it is weaving assistants and automation into shopping, coding, and business software. (aboutamazon.com, cnbc.com) Jassy made the comparison as blunt as possible in his letter: three years after Amazon Web Services launched commercially, it was running at $58 million a year, while three years into this artificial intelligence wave, Amazon Web Services artificial intelligence is already above $15 billion. He used that contrast to say this market is arriving much faster than cloud computing did. (aboutamazon.com) The bet is that artificial intelligence will not stay a premium product for a few labs and start-ups. Amazon is spending as if cheaper chips, bigger data centers, and built-in tools will turn artificial intelligence into a standard utility that companies buy the way they already buy storage, databases, and computing power. (aboutamazon.com, reuters.com) What changed this week is that Amazon stopped asking investors to imagine that future. It said the business has already reached a $15 billion annual pace, and then it argued that the next fight is not whether artificial intelligence will be a business, but who can afford to build enough infrastructure to keep up. (reuters.com, aboutamazon.com)