• Oracle (ORCL), Alphabet (GOOGL), Microsoft (MSFT) and Amazon (AMZN) all traded lower as investors reassessed AI spending and financing risks.
  • The selloff extended beyond individual names, with the S&P 500 down 0.8% and Nasdaq Composite down 1.3% in early trading.
  • Meta’s earnings intensified scrutiny of debt-funded AI expansion, while higher oil prices and inflation concerns added to the pressure.

A Broad AI Repricing

Shares of major U.S. hyperscalers fell on Tuesday, tracking a wider decline in AI-related stocks as investors questioned whether rapidly growing cloud revenues can justify enormous data-center spending, rising financing costs and increasingly complex debt commitments. Oracle dropped 2.4%, Alphabet fell 2.3%, Microsoft slipped 1.5% and Amazon declined 4%, according to intraday trading data. The moves came as the S&P 500 was down 0.8% and the Nasdaq Composite 1.3% at 9:35 a.m. Eastern, following larger declines in Asian markets with substantial technology exposure.

The selloff was not confined to the four companies. Meta Platforms (META) fell roughly 6% after its earnings report showed higher capital-spending plans, renewing concerns about debt-funded AI expansion. Alphabet, Microsoft and Amazon are scheduled to report the following week, making their spending guidance and cash-flow outlook immediate catalysts for the market.

Financing Concerns Mount

Apprehension about AI-related liabilities had been building before Tuesday’s headline. On October 19, S&P Global (SPGI) warned that major technology companies could have more AI-linked liabilities than publicly apparent because of interconnected financing arrangements, though the rating agency saw no immediate need to change ratings. The warning followed an October 14 AI selloff that was triggered by a hot U.S. inflation report: CoreWeave (CRWV) dropped 16%, Nebius (NBIS) 11%, Oracle 9% and Dell (DELL) 11%, illustrating the sensitivity of infrastructure-heavy businesses to borrowing costs.

“What institutional investors like us are really focused on is regulatory stability,” said one market participant, speaking on condition of anonymity. “But when financing costs rise, the calculus for these massive buildouts changes quickly.”

Spending vs. Cash Flow

The central issue is whether AI revenues will justify the investment required to supply demand. A July Reuters (TRI) analysis found that consensus estimates for the five major hyperscalers’ annual capital expenditure had risen from approximately $485 billion in January to $730 billion in July. Between 2025 and 2027, projected annual operating cash flow increased by about $340 billion, while capital spending increased by roughly $534 billion.

The financial exposure is uneven. Oracle is more funding-sensitive: in fiscal 2026 it spent $55.7 billion on capital expenditure against $32 billion of operating cash flow. Amazon raised its 2026 capital-spending forecast to $220 billion and said capacity would still be insufficient to satisfy demand in 2026 and likely 2027. Microsoft, Alphabet and Meta still generated enough free cash flow to cover dividends and buybacks in their latest fiscal years, providing a cushion that distinguishes them from more financing-dependent operators.

Falling share prices do not, by themselves, demonstrate collapsing AI demand. Recent company results show strong growth: Oracle’s fiscal Q1 2027 revenue rose 30% to $19.35 billion, with cloud revenue up 62%; Alphabet’s Q2 2026 revenue increased 24% to $119.8 billion; Microsoft’s fiscal Q4 2026 revenue grew 18% to $90.0 billion; and Amazon’s Q2 2026 revenue rose 20% to $200.6 billion. Investors are reassessing the price and financing required to supply that demand, not the demand itself.

Geopolitical and Macro Pressures

The technology-sector weakness coincided with higher oil prices and concerns about the war with Iran, extending the selloff beyond individual companies. Rising oil prices add inflation uncertainty, making data centers more expensive to fund. Because hyperscalers are major buyers of chips, servers and construction services, doubts about their budgets can spread to suppliers and infrastructure lenders.

Technology controls are another backdrop. On October 16, China proposed adding 40 technologies—including AI models, agents and computing clusters—to its restricted or prohibited technology-export catalogue, while introducing interim pre-export registration requirements. Consultation runs until November 15. The proposal could complicate cross-border technology deployment, though the reviewed evidence does not establish it as a direct cause of Tuesday’s decline.

What to Watch

The next major tests are earnings from Alphabet, Microsoft and Amazon. Investors are likely to focus on cloud growth, capital-spending revisions, financing requirements and the timing of cash-flow recovery. Meta’s reaction shows that encouraging revenue alone may not satisfy the market when investment commitments rise.

Oracle has also undergone substantial workforce restructuring, with fiscal-2026 job cuts at roughly 21,000 while the company accelerated AI infrastructure investment—a contrast that highlights the redistribution of resources toward cloud capacity. Clay Magouyrk and Mike Sicilia became co-CEOs on September 22, 2025, a transition that predates the current selloff.

Two outcomes remain plausible: strong demand converts into adequate returns, or investment outruns profitable demand and weighs on cash flow and valuations. S&P’s digital-infrastructure head, Mark Habib, identified profits failing to materialize as the principal risk to hyperscaler ratings—not simply the existence of large spending plans. For now, the evidence supports a cautious interpretation: this is a reassessment of AI economics and financing, not proof that the underlying technology market has stopped growing.

Correction: An earlier version of this article misstated the timing of Oracle’s co-CEO transition. It occurred on September 22, 2025.