This is how Microsoft is making money from AI right now.

Investors are increasingly concerned about whether the massive amounts of money being spent on artificial intelligence (AI) infrastructure will ever deliver returns. While tech companies argue that AI is a must-invest area and that underspending poses a greater risk than overspending, data center and cloud providers are still committing hundreds of billions of dollars to buildouts. Investors are unlikely to tolerate such capital expenditures indefinitely without evidence of payoff.

How Microsoft is monetizing AI

Some AI hyperscalers are already generating revenue tied to AI demand, and there is a pathway to expanding that income. In Microsoft’s case, the company’s AI-related revenue is primarily driven by two areas: Copilot and Azure.

Copilot: AI embedded across Microsoft products

Microsoft is integrating AI throughout its products, with Copilot as a central offering. Copilot is embedded into nearly every Microsoft Office product, and subscribers pay extra to access the add-on. The article notes that this has helped Microsoft boost revenues from its core software offerings.

Azure: cloud growth tied to AI demand

Azure, Microsoft’s cloud computing division, is described as the more significant driver of AI-related growth. The article states that Azure revenue increased by 39% year over year in the fourth quarter. It also notes that the growth rate would have been even higher if Microsoft had leased more of its new computing capacity to external customers rather than dedicating it to internal use.

The underlying model is that Microsoft builds excess computing capacity and rents it to customers that need it. The article emphasizes that many companies lack the resources and expertise to build their own AI data centers, making cloud services an attractive alternative. It also adds that many clients prefer not to carry data center assets on their balance sheets.

Azure is characterized as a usage-based or subscription-based service, meaning Microsoft earns more as customers use more capacity. As AI demand rises, the article suggests, Azure revenue can rise as well.

Capacity expansion and AI computing backlog

Microsoft is expanding its capacity to meet a large AI computing backlog. The article cites a $625 billion backlog for AI computing as a key factor driving additional spending on AI infrastructure, arguing that what is already online is not close to what would be needed to support an AI-first economy.

Implications for investors

The article concludes that, over the long term, the revenue generated from Azure is expected to far outpace what Microsoft is spending. It frames this as a reason to view Microsoft’s AI investment as financially sound, with the stock described as a smart investment for investors who focus on the long term.