Microsoft, Amazon results prove AI capex is paying off

Microsoft and Amazon have posted quarterly results that signal a turning point for the artificial intelligence sector, with both companies reporting cloud growth constrained only by hardware availability rather than customer interest. Azure revenue grew 39% year-on-year, pushing total Microsoft Cloud revenue past $51.5 billion. Amazon Web Services expanded 37% to $42.2 billion, its fastest pace in over four years.

Executives at both firms highlighted that demand for AI computing now vastly exceeds their ability to supply it. Microsoft attributed Azure's growth limits to a shortage of graphics processing units. Amazon warned that demand could outstrip available cloud capacity through 2028. Jefferies data illustrates this bottleneck, projecting global data centre demand will hit 21.1 GW in 2025 against operational capacity of just 8.9 GW.

In response, Amazon has raised its planned 2026 capital expenditure to nearly $220 billion. The market rewarded the outlook aggressively. Microsoft shares jumped more than 15%, while Amazon climbed 15.3% after reporting over $62 billion in quarterly profit and 20% overall revenue growth.

For nearly two years, investors worried that heavy spending on AI infrastructure would crush profit margins without generating returns. These results suggest the investment cycle is finally yielding commercial gains. "Some investors have questioned whether the spending on AI infrastructure would translate into stronger earnings," said Ross Maxwell, Global Strategy Operations Lead at VT Market. "The latest results suggest that we are beginning to see that come to fruition, with strong revenue growth supporting the significant capital investment."

Analysts view the reports as evidence the industry has entered a new phase. Ponmudi R, CEO of Enrich Money, noted that the sector has moved past initial excitement and pure infrastructure building into a stage where those investments produce measurable revenue. "The biggest takeaway from these earnings is simple: AI demand is no longer the question," Ponmudi said. "The real challenge is building enough infrastructure to keep up with it."

The shift carries clear implications for equity valuations. Maxwell noted that while Microsoft and Amazon have diverse business models capable of sustaining AI investment without damaging profitability, the bar is now rising for the rest of the sector. Companies that fail to convert heavy AI spending into top-line growth will struggle to justify their premium valuations.