Make this analysis yours. Add it to a free WorldbyFlow workbench to run follow-ups, ask questions, and re-check it as events move.

By the Numbers

Four hyperscalers (Microsoft, Amazon, Alphabet, Meta) guided to roughly $725 billion in combined 2026 capex, up 77% from ~$410 billion in 2025 — while MIT-affiliated research finds only 11% of S&P 500 firms have deeply integrated AI into production operations as of 2025.

This scan sets hyperscaler AI infrastructure spending commitments — running toward roughly $1 trillion-plus annually by 2027 — against hard data on how much of that capital is translating into production-grade enterprise AI deployment, where most large companies remain stuck in pilot or narrow-function use. The gap between capital committed and value realized is now the central quantitative question in the AI investment cycle.

The capital side of this story is well-documented and largely consistent across sources. Google, Amazon, Microsoft, and Meta collectively plan to spend $725 billion on capex in 2026, up 77% from last year's record $410 billion, according to first-quarter earnings compiled by the Financial Times. Company-level guidance underlying that total: Amazon leads AI infrastructure spending at $200B in 2026, ahead of Google ($185B), Meta ($125B), and Microsoft ($120B), with $725B total, up 77% from $410B in 2025. Looking further out, America's biggest technology companies plan to spend $900bn on artificial intelligence infrastructure in 2025, rising to $1.4trn by 2027 in history's largest investment surge per the Economist's July 2026 analysis — a materially higher trajectory than the earnings-based $725B/2026 figure, reflecting different scoping (global vs. four US hyperscalers) and methodology. Separately, Goldman Sachs analysis cited by Yahoo Finance projects Alphabet's $175-185 billion, Amazon's $200 billion, and Meta's $115-135 billion capex guidance, which dramatically exceeded consensus estimates, and Goldman now expects a combined $5.3 trillion of capex spending for the four largest hyperscalers from fiscal year 2025 to fiscal year 2030, up from a prior estimate of $4.5 trillion — a projection, not a committed figure. The spending-to-revenue mismatch is stark: capex is reaching seemingly untenable levels as a share of sales — roughly 86% for Oracle, 54% for Meta, 47% for Microsoft, 46% for Alphabet, and 25% for Amazon in 2026, and independent analysis finds the four US hyperscalers purchased $433.9 billion of property and equipment in the four quarters through March 2026 against roughly $149 billion of reported depreciation over the same span, a structural capex-to-depreciation gap that means today's income statements understate the true scale of the build-out.

On the deployment side, the data diverges sharply depending on definition and measurement rigor — and this divergence is the central analytical finding. The most legally rigorous benchmark comes from MIT-led research using 10-K disclosures: deep AI integration stood at 5% of S&P 500 firms in 2022; by 2025 that figure had more than quadrupled to 11%, with another 10% of firms using AI at a production or service-delivery level below full integration. That same research documents a profitability J-curve and notes the researchers found no statistically measurable difference in capital expenditure or productivity between adopters and non-adopters at this stage, suggesting the financial signal of AI investment has not yet reached the income statement or balance sheet for most firms. The Census Bureau's Business Trends and Outlook Survey, a biweekly nationally representative instrument, tells a comparable but distinct story using a broader adoption definition: AI use among US businesses remained between 17% and 20% from December 2025 to May 2026, with around 37% of businesses with at least 250 employees reporting AI use in operations, versus 32% for firms with 100-249 employees, as of the period ending May 3, 2026. A Census working paper narrows the lens further: during the supplement reference period of November 2025 to January 2026, 18% of firms used AI in a business function, rising to 32% on an employment-weighted basis; adoption is expected to reach 22% within six months. Crucially, breadth of adoption does not equal depth of integration — among businesses that report current AI adoption, 57% apply AI tools in three or fewer operational functions, per an analysis of the same Census dataset.

The pilot-to-production failure rate is the most widely cited — and most contested — figure in this landscape. MIT's Project NANDA, in a report released in July 2025, found that despite $30-40 billion in enterprise investment, 95% of generative AI projects yield no measurable business return, and 95% of organizations are seeing no business return from generative AI spending. The sharpest form of that finding: only 5% of custom enterprise AI tools reach production, with enterprises specifically lagging — enterprises run the most pilots but convert the fewest; mid-market organizations move faster from pilot to full implementation (~90 days) than large enterprises (nine months or longer). This MIT figure should be flagged for its vintage: it is a July 2025 report being recirculated through 2026 commentary, and its methodology rests on 52 executive interviews, surveys of 153 leaders, and analysis of 300 public AI deployments (other citations of the same report describe slightly different interview/survey counts — 150 interviews and 350 employee surveys in some tellings — a discrepancy that should itself be treated as a data-quality caveat). A more recent industry survey offers a partially contradicting signal on outcomes even as failure-to-scale metrics hold: a Domino Data Lab report released July 21, 2026 found that the share of enterprises whose ROI fails to outpace their investment has held at 57% since 2025, even as enterprise AI production capability continues to climb — suggesting production deployment is rising in absolute terms even as ROI conversion stays flat. Consultancy-sourced figures reinforce the same pattern from a different angle: Deloitte's survey of 3,235 leaders across 24 countries found 66% report productivity gains, but only 20% see AI-driven revenue growth, and only 6% of companies convert spending into enterprise-wide financial impact despite 86% of budgets rising.

The Federal Reserve's own framing of this gap, published July 17, 2026, is the most institutionally authoritative synthesis available and explicitly names the divergence this scan is measuring: the current gap between financial markets, which have been highly responsive to changes in the AI narrative, and aggregate output and labor market data, which currently show limited signs of broad-based transformation is the object of ongoing Fed monitoring. The Fed's own historical Census-based citation adds texture: micro-level productivity gains are documented, but the broader question is whether these gains translate into aggregate productivity growth. Agentic AI, the leading edge of production deployment, shows the same divide in miniature: the gap between 80% of enterprise apps embedding an agent and 31% running one in production is the real 2026 story — embedding is easy, operating is hard, per one industry compilation citing Gartner and McKinsey research.

The claims behind this analysis, each with its verification status — including what is contested, unverified, or could not be established.

August 7, 2026

August 3, 2026

July 28, 2026