Microsoft, Amazon and Google could spend nearly $600 billion on AI infrastructure in 2026, as investors look for evidence that cloud growth can justify the bill. The near-term test arrives with earnings from the three companies due on 28 and 29 October.
Watch Desk analysis
What happened
Coinpaper reports that the companies’ combined contracted cloud backlog is approaching $1.7 trillion. It cites analyst forecasts for September-quarter cloud growth of 95.1 per cent at Google Cloud, 47.7 per cent at Microsoft Azure and 38.2 per cent at Amazon Web Services. Those are forecasts, not reported results.
The article puts planned 2026 spending at about $220 billion for Amazon, $195 billion to $205 billion for Alphabet, Google’s parent, and roughly $175 billion for Microsoft, including finance leases. The figures underline the scale of the infrastructure build-out: data centres, chips, power and networks must be paid for before the hoped-for returns are fully in.
Why it matters
Cloud growth is strong, but growth alone does not settle whether the AI investment pays. Investors will also be watching margins and cash flow, particularly as companies expand capacity to meet demand. Coinpaper says Amazon’s spending has already put pressure on its cash flow, despite recent AWS growth.
The timing makes this more than a very large number doing laps around the news cycle. These earnings will offer fresh evidence about demand and the costs of serving it, though they cannot by themselves resolve whether the wider AI build-out will earn an adequate return.
Our read
The useful question is not whether these firms are spending heavily. They plainly are. It is whether cloud revenue, profitability and cash generation are keeping pace with the infrastructure required to deliver AI services. Watch the results, but keep forecasts separate from outcomes; even a trillion-dollar backlog is not the same thing as profit in the bank.
What to watch
- Whether Google Cloud, Azure and AWS meet or exceed the cited growth forecasts.
- What the companies say about cloud margins, cash flow and infrastructure costs.
- Whether earnings disclosures change their spending plans or clarify expected returns.
Discussion spark: Should investors judge AI infrastructure spending mainly by cloud growth, or insist on clear evidence of margins and cash returns before giving the build-out more credit?
Sources and evidence
- Source update (27 October 2026, 12:37 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.