Microsoft and Meta have reported another quarter of rapid revenue growth linked to artificial intelligence, but the two sets of results showed how differently the cost of the AI race is flowing through Big Tech balance sheets.

Microsoft converts demand into cloud growth

Microsoft reported quarterly revenue of $90.0 billion, up 18% from a year earlier. Operating income rose 18% to $40.6 billion and GAAP net income increased 31% to $35.8 billion.

The strongest signal came from Azure and other cloud services, where revenue grew 43%. Microsoft said annual Azure revenue passed $100 billion for the first time, while Microsoft Cloud revenue reached $59.3 billion in the quarter, an increase of 27%.

The figures suggest that Microsoft is converting heavy spending on data centres, chips and model access into large-scale commercial demand. Businesses are using Azure for conventional computing as well as AI training, inference and software services, giving Microsoft several routes to recover infrastructure costs.

Microsoft also reported more than 30 million paid Microsoft 365 Copilot seats. That figure is important because it measures the company's attempt to turn generative AI into recurring software revenue rather than treating it only as a feature used to defend existing products.

Meta grows revenue but absorbs higher costs

Meta also delivered strong top-line growth. Revenue rose 28% to $60.8 billion, supported by higher advertising volumes and prices across its family of apps.

The cost side was much less favourable. Total costs and expenses increased 55% to $42.0 billion. Operating income fell 8% to $18.8 billion and net income declined 14% to $15.8 billion.

Meta spent $31.1 billion on capital expenditure and finance-lease principal payments during the quarter. It now expects full-year 2026 capital expenditure of between $130 billion and $145 billion, reflecting the enormous infrastructure requirements of its AI strategy.

Not all of the increase was caused by data centres or chips. Meta said the quarter included $2.4 billion in legal charges and $1.18 billion in severance costs connected to its May workforce reduction. Even so, the scale of planned capital spending means investors will continue to test whether AI improvements generate enough advertising, engagement and new business to justify the outlay.

Two models of the same investment cycle

Both companies are spending heavily, but they enter the cycle from different positions. Microsoft sells cloud capacity, software subscriptions and developer tools directly to companies. That allows infrastructure demand to appear quickly as reported revenue.

Meta uses AI primarily to improve advertising, recommendations and products used by consumers. Those gains can be valuable, but the link between a new data centre and an additional dollar of profit is less direct. Meta is also investing in longer-term products that may take years to produce significant revenue.

What the results say about the AI boom

The earnings do not show that one strategy has succeeded and the other failed. Meta still expects third-quarter revenue of $61 billion to $64 billion, and its advertising business remains highly profitable. Microsoft also faces rising infrastructure costs and must keep expanding capacity to meet demand.

They do show that the market is becoming more selective. Companies are no longer judged only on whether they are spending on AI. The harder question is how quickly that spending becomes durable revenue, margin and cash flow.