- Published on
Investors love AI, as long as you’re a cloud host
- Authors
- Name
- DeepHorus
Amazon reported better-than-expected second-quarter earnings on Thursday, with net sales rising 20% year over year. Cloud revenue stood out as a particular bright spot, driving the company’s stock up nearly 10% in after-hours trading. The results underscore investors’ growing enthusiasm for cloud hosts that can monetize artificial intelligence, even as spending on AI infrastructure balloons.
The tech giant’s earnings come amid a broader AI investment boom, where companies are pouring billions into data centers and chips. Amazon is not slowing down its capital expenditures despite conventional wisdom that investors want restraint. The company spent $173 billion on property and equipment for the fiscal year ended June 30, up from $107.65 billion a year earlier, covering GPUs, natural gas turbines, and land.
Amazon also raised its 2026 capex forecast from $200 billion to $220 billion, even as it dipped into cash reserves. The company ended the quarter with $7.6 billion less cash than 12 months ago, marking its first negative free cash flow period this year. However, AWS revenue rose 37% year over year to $42 billion for the quarter, demonstrating that demand is growing alongside supply.
The robust cloud revenue helps justify the massive spending, especially given the time lag between data center construction and capacity sales. Amazon is also making long-term bets on custom chips like Trainium and the Arm-based Graviton processor, which could improve cloud margins. CEO Andy Jassy noted during the earnings call that the AI business follows a similar margin trajectory to the core business, emphasizing that AWS doesn’t need its own frontier model to succeed.
This dynamic is not unique to Amazon. Microsoft and Google also saw stock gains after reporting strong cloud revenue. In contrast, Meta faced intense skepticism, with its stock falling 8% after earnings, as investors focused on its high capex and lack of clear AI revenue. The pattern reveals that investors currently favor cloud hosts as the most reliable part of the AI stack, while remaining wary of AI labs and startups.
Yet, the sustainability of this demand remains a critical question. Amazon’s hosting revenue is effectively someone else’s AI bill—for instance, Anthropic’s spending flows to AWS. If AI labs and their clients cannot sustain their spending, cloud hosts could face a downturn. The ultimate question, as analyst David Cahn posed, is whether there is enough demand to justify the buildout.
While cloud hosts like AWS are a few steps removed from the demand problem, they are not insulated from it. The coming years will test whether the massive infrastructure investments pay off or lead to a shakeout. For now, investors are betting on the hosts, but the AI economy’s future hinges on real, enduring demand.
