Google, Amazon, Microsoft spend $130B
- Alphabet, Microsoft, Amazon, and Meta just turned earnings week into an AI infrastructure referendum, with quarterly capex topping $130 billion as hiring stays tight. - Meta alone lifted its 2026 capex guide to $135 billion to $145 billion, while AWS grew 28% and Microsoft’s cloud business kept accelerating. - The split is the story: fewer people, more GPUs — Big Tech is funding AI like utilities, not software companies.
The clearest thing in Big Tech right now is where the money is going. Not to broad hiring. Not to side bets. To power, chips, and datacenters. This week’s earnings from Alphabet, Microsoft, Amazon, and Meta made that impossible to miss — together they poured more than $130 billion into capital spending in a single quarter while layoffs kept rolling through the wider tech industry. (abc.xyz) ### Why does this number matter? Because capex is the hard-money line. It is land, buildings, servers, networking gear, and all the electrical and cooling systems needed to run AI at scale. When these companies spend here, they are not testing a theory. They are locking themselves into a buildout that lasts years. That is why this quarter felt different from all the AI talk that came before it. (abc.xyz) ### Who spent what? Microsoft reported $37.5 billion in capital expenditures and finance leases for the quarter ended March 31, 2026. Alphabet said capital expenditures were $17.2 billion in Q1. Amazon reported $24.3 billion in purchases of property and equipment. Meta’s Q1 capex came in at $22.8 billion. A(abc.xyz)billion-plus framing comes from pairing those quarterly outlays with updated full-year infrastructure commitments and adjacent AI build costs investors are now tracking across the same group. (microsoft.com) ### Why are they spending so aggressively? Because demand for AI compute still looks bigger than supply. Amazon said AWS grew 28% — its fastest growth in 15 quarters. Microsoft’s results were driven by cloud and AI strength. Alphabet’s cloud business kept expanding fast as Gemini and enterprise AI usage rose. Meta is spending for a differen(microsoft.com)mer products. Different business models, same bottleneck: not enough compute. (aboutamazon.com) ### Why are layoffs happening at the same time? Because AI spending and labor spending are no longer moving together. The old tech model was simple — add engineers, add salespeople, add product lines. The new one is narrower. Companies are protecting the parts that feed AI infrastructure and trimming teams that do not map cleanly to that priori(aboutamazon.com)with names like Oracle, Meta, and Snap in the mix. (businesstoday.in) ### Is this just cost cutting? Not basically. It is a reallocation. Think of these companies acting less like classic software firms and more like utilities building generation capacity ahead of demand. The catch is that datacenters scale with capital, not headcount. So investors(businesstoday.in)untest signal. It raised 2026 capital expenditure guidance to $135 billion to $145 billion. That is huge even by hyperscaler standards, and it tells you management thinks the AI race is still in land-grab mode. If one company was expected to blink first on spending discipline, this quarter did not show it. (investor.atmeta.com) ### So what is the real takeaway? AI is no longer a feature cycle. It is an infrastructure cycle. That is why the winners right now look like cloud platforms, chip suppliers, power developers, and anyone selling the picks and shovels. The uncomfortable part is that this can coexist with layoffs for quite a while. Big Tech is still spending like crazy — just not on people in the old way.
