Alphabet Slips 2% Despite NASDAQ Rally as Pichai Says Google Cloud Is Giving Up Revenue for AI Capacity; Amazon Dips, Microsoft Holds Steady

Alphabet shares fell 2% to $331.83 while the Nasdaq 100 rose 1% after CEO Sundar Pichai said Google Cloud is forgoing short-term revenue to reserve computing capacity for frontier AI and Google DeepMind. He said serving new customers seeking capacity for next year would be “incredibly difficult,” but did not quantify foregone revenue or provide a timeline for added capacity. Investors are looking to Alphabet’s upcoming Q3 2026 report for clarity; Microsoft fell 0.3% and Amazon 0.5%.

Analysis

The key risk is not simply foregone near-term cloud sales: allocating scarce compute to DeepMind could defer customer workloads long enough for them to establish on AWS or Azure, raising future switching costs and weakening Google Cloud’s ability to convert demand into durable share. Conversely, internal access to frontier compute may improve Google’s models and products, creating indirect value across cloud and advertising; the interview provides no evidence yet to quantify that payoff. The near-term market problem is therefore uncertainty around the opportunity cost, not proof that the trade-off is financially material.

Over the next 1–3 months, the next report should clarify whether cloud growth, backlog/commitments, and capacity investment show that demand is being deferred or permanently lost. Watch for management to distinguish capacity-constrained demand from weaker bookings, and for any indication of when new capacity becomes available to external customers. Over 6–18 months, power, cooling, and networking bottlenecks could constrain all major cloud providers, limiting the share-transfer benefit to AWS and Azure while supporting infrastructure suppliers. The contrarian case is that investors may be treating an unquantified, potentially temporary constraint as lost revenue without crediting the strategic value of internal AI capacity; that case strengthens only if cloud growth and customer commitments remain robust.

Key Decisions for Investors

• Avoid adding to GOOG solely on the assumption that this is a temporary capacity issue. Before the next report, monitor cloud growth, backlog or remaining performance obligations, capex/capacity commentary, and evidence of customer deferrals; the article supplies no dollar estimate of revenue forgone.
• For a near-term relative-value expression, consider a modest underweight GOOG versus a diversified mega-cap or cloud peer basket rather than an outright short: the capacity-allocation uncertainty is company-specific, but AWS and Azure may also face industry-wide infrastructure limits. Reassess after management quantifies the constraint.
• Treat AMZN and MSFT as potential beneficiaries only if displaced Google Cloud workloads are actually being captured; seek evidence in cloud growth, bookings, or capacity commentary before positioning on share gains. Do not infer a direct earnings benefit from Google’s comments alone.
• Falsification: reduce the relative underweight if GOOG reports resilient cloud growth and commitments alongside a credible timeline for external capacity expansion, or demonstrates measurable AI-related product benefits. A worsening growth trend or evidence that customer demand is shifting to rivals would reinforce it.