AI investing enters a new phase as markets demand results

AI investing is entering a more selective phase in which execution and measurable financial returns matter more than AI-related excitement.

• Microsoft shares rose 15% after Azure growth accelerated to 43%, annual Azure revenue exceeded US$100 billion, and Microsoft 365 Copilot surpassed 30 million paid seats.
• Microsoft signed more than US$130 billion of data-centre leases in the June quarter, bringing future commitments to US$329 billion, while Amazon Web Services grew 37% to US$42.2 billion and Amazon raised planned capital spending to about US$220 billion.
• Alphabet reported revenue growth of 24% to US$119.8 billion, Google Cloud growth of 82% to US$24.8 billion, a US$514 billion cloud backlog and 950 million monthly Gemini users, but raised 2026 capital-expenditure guidance to as much as US$205 billion.
• Meta’s Reality Labs lost US$4.6 billion in the latest quarter, with cumulative operating losses above US$80 billion since late 2020, illustrating investor concern about the cost and payback of AI ambitions.
• Apple’s revenue rose 16% to more than US$109 billion, while iPhone and Mac sales increased 22% and 29%; however, it forecast 9%–11% quarterly sales growth because memory and chip shortages constrained supply.
• Semiconductor sales have risen about 175% in the current AI cycle, below the roughly seven-fold increase during the late-1990s internet build-out, while earnings growth rather than expanding valuation multiples has driven much of the sector’s share-price gains.

Takeaway: AI remains a major long-term investment theme, but investors should favour companies that can convert spending into revenue, earnings, cash flow and durable advantages, while treating large capital commitments without visible payback with greater caution.