Microsoft: AI Demand Is Real. The Bill Is Bigger.
Microsoft’s (MSFT) latest numbers say the quiet part out loud. AI is driving growth. It is also driving record spending. Your question is simple. Do margins hold as scale kicks in, or do returns fade as the bill grows.
Start with results. For the quarter ended June 30, 2025, revenue reached $76.4 billion, up 18 percent. Operating income hit $34.3 billion, up 23 percent. The Intelligent Cloud segment delivered $29.9 billion, up 26 percent. These figures frame the story. AI is pushing Azure and the broader cloud stack higher.
Management tied more of Azure’s growth to AI each quarter through fiscal 2025. In the March quarter, AI services added 16 percentage points to Azure’s 33 percent growth. The mix shift toward AI workloads keeps accelerating consumption and encourages longer commitments. For investors, this improves visibility.
Copilot gives the clearest proof on the application side. GitHub Copilot crossed 20 million all-time users by late July. Microsoft also reported the largest quarter of Copilot seat adds since launch. Price remains firm at $30 per user per month for Microsoft 365 Copilot. Monetization, not novelty, drives ROI on the AI buildout. These details matter for your model.
Now the bill. AI capex moved from large to historic. Fiscal 2025 capex totaled about $88.7 billion. Fourth-quarter capex was $24.2 billion. Management guided to roughly $30 billion for the September quarter. The driver is supply. You build data centers and secure accelerators to meet backlog. This front-loads depreciation and power costs. It also sets a higher base for free cash flow to clear.
The spend shows up in new footprints. Microsoft announced over $30 billion of investment in the UK over four years, including thousands of advanced AI chips. In the United States, it lifted planned spending in Wisconsin to more than $7 billion to add a second AI data center. These moves reduce supply bottlenecks and pull demand into Microsoft’s regions. They also harden the moat around Azure.
How to frame margins. In the near term, gross margin mix tilts lower because AI services carry higher cost of revenue. Utilization and pricing improve as new capacity fills. Watch three signals. First, Azure growth relative to reported AI contribution. Second, software pull-through from Copilot across Microsoft 365, Dynamics, and Security. Third, opex discipline while capex peaks. Microsoft’s breakdown shows FY2025 revenue up 15 percent year over year with cost of revenue up 19 percent, driven by cloud scale. The spread should narrow as cohorts mature.
For Copilot, adoption is broadening beyond pilots. Large enterprises are rolling out tens of thousands of seats. The standard $30 price point gives a clean way to model ARPU uplift in Microsoft 365. Independent studies and Microsoft case data help finance teams quantify time savings. That keeps procurement onside and supports renewal rates.
Risks deserve attention. AI supply cycles cut both ways. If demand slows or power constraints persist, new capacity weighs on returns. Competitive pressure remains intense in models and AI infrastructure. A weaker macro backdrop would also delay Copilot seat expansion or limit usage intensity. Treat these as standing risks until margin expansion proves durable.
What to watch next
• Azure growth with explicit AI contribution. Look for AI to add double-digit percentage points to Azure growth again.
• Copilot monetization. Seat adds, price integrity at $30, and reference wins in regulated industries.
• Capex path. Whether quarterly spend holds near $30 billion or starts to roll down as supply catches up. This drives the free cash flow inflection.
My take? The setup is binary enough for a mega cap. If Microsoft keeps translating AI demand into Azure growth and Copilot ARPU, the stock earns its premiums. If the spend outruns monetization, estimates face pressure. For now, the reported numbers support the bull case. Keep your eyes on growth mix, price discipline, and the capex slope.