Morgan Stanley flags surging AI data center spending as it lifts Amazon and Meta capex outlook

Analysts at Morgan Stanley on Monday raised their capital expenditure estimates for major cloud and AI infrastructure spenders, pointing to continued upward pressure from the cost of building and operating AI data centers.

The market discussion, reported by Yahoo Finance, said AI-focused data center spending by cloud “hyperscalers” could reach $1.4 trillion by 2028. Within that broader capex push, Morgan Stanley increased estimates for Meta Platforms and Amazon, indicating investors may need to brace for continued heavy spending even as AI demand remains a central growth narrative.

For Amazon, the capex lift reflects how AWS and related operations are positioned to benefit from AI workloads that require specialized compute, power, cooling, and networking. In plain terms, the more companies rely on cloud-hosted AI systems, the more they depend on data center build-outs and upgrades that can be measured in capital spending.

Amazon’s disclosure practices typically separate large spending categories across its businesses, with AWS and fulfillment and transportation costs tracked through its financial reporting. However, the Yahoo Finance item did not provide additional company-level details about the specific projects or timing behind the revised estimates, nor did it quote Amazon management on incremental AI infrastructure plans.

Meta’s inclusion underscores that the data center spending surge is not limited to one business model. Meta is also a major buyer of advanced computing capacity for AI initiatives, ranging from internal ranking and recommendation systems to generative AI products. Morgan Stanley’s combined framing suggests analysts view AI infrastructure demand as broad-based across multiple hyperscalers.

Morgan Stanley’s raised figures arrive at a time when investors have focused heavily on how AI changes the cost structure of the internet economy, especially the balance between near-term capex and longer-term revenue generation. If spending rises faster than monetization, the market can treat earnings and free-cash-flow timing as a near-term risk, even if revenue growth remains strong.

For Amazon specifically, the company’s official newsroom and ongoing updates tend to focus on service expansion and operational milestones, rather than giving a forward view of exact capex revisions made by sell-side analysts. As a result, Tuesday’s market narrative points to investor-facing estimates rather than an Amazon announcement of higher spending. The Yahoo Finance report did not spell out how much higher Amazon capex would be under Morgan Stanley’s updated framework, nor did it disclose the bank’s detailed assumptions.

What to watch next is whether Amazon and other hyperscalers provide clearer indicates in upcoming financial materials about the pace of AI infrastructure investments. Investors will likely look for any guidance or commentary that ties capex to measurable demand drivers, such as customer commitments for AI capacity or signs that utilization of new hardware is climbing enough to offset spending over time.