AI Spending Hits $730B But Only 12% of CEOs Can Show Real Returns

Quick Facts

• The four major hyperscalers are on track to spend approximately $730 billion on AI infrastructure in 2026, up 78% from 2025.
• Only 12% of CEOs report AI has delivered both cost savings and revenue gains, per a PwC survey of 4,454 executives across 95 countries.
• MIT found 95% of AI pilots deliver zero measurable P&L impact, while 72% of CIOs are breaking even or losing money on AI deployments, per Gartner.

The central question hanging over the AI boom is not whether the technology works. It is whether anyone can prove it on a balance sheet. The gap between AI spending and verified AI returns has become the defining tension of 2026.

The scale of investment is historically unprecedented. Amazon, Google, Microsoft, and Meta are collectively targeting roughly $730 billion in AI capital expenditure this year, according to company-by-company capex guidance from Q3 2026 earnings calls. That figure is up 78% from approximately $410 billion in 2025. Amazon leads at around $220 billion, followed by Google at $195 to $205 billion, Microsoft at roughly $170 billion, and Meta at $135 to $145 billion.

Collectively, Amazon, Alphabet, and Microsoft are spending 102% of their cloud revenue on capital expenditure in 2026, effectively recycling all cloud income back into AI infrastructure. UBS projects the three will spend $4.1 trillion on AI infrastructure from 2026 through 2028.

The Return Problem

The spending has not produced proportional proof of returns. PwC's 29th Global CEO Survey, published in January 2026 and covering 4,454 chief executives across 95 countries, found that only 12% of CEOs reported AI delivered both cost savings and revenue benefits over the prior 12 months. Fifty-six percent reported neither outcome.

BCG and KPMG surveys of more than 2,100 executives found that just 5% to 8% of enterprises report measurable AI ROI despite average AI budgets of $186 million and 88% adoption rates. KPMG's mid-2026 AI Pulse survey of 2,145 leaders put the share that had established ROI at 7%.

McKinsey's State of AI 2026 report found that 80% of respondents say AI has improved their individual productivity. Yet only 37% attribute any EBIT impact to AI. Just 6% qualify as high performers, defined as organizations attributing at least 5% of EBIT to AI.

Failure Rates Across the Industry

The failure data is consistent across multiple research organizations. MIT found that 95% of AI pilots deliver zero measurable P&L impact. S&P Global found that 42% of companies abandoned most of their AI projects in 2025, up from 17% the year before. IBM put the share of initiatives delivering expected ROI at 25%. Morgan Stanley found only 21% of S&P 500 companies could cite a measurable AI benefit.

Deloitte's 2025 survey found that only 6% of respondents achieved ROI from a typical AI use case within one year.

The Aggregation Paradox

Researchers have identified a structural reason why individual productivity gains do not show up in company financials. The International Labour Organization has named it the aggregation paradox. AI delivers large productivity gains at the task level, often 10% to 70%, but those gains do not translate to firm-level or macroeconomic productivity growth.

The math is straightforward: tasks are only a fraction of a job, jobs are only a fraction of a firm, and AI-exposed firms are only a fraction of the economy. In a randomized experiment with 758 consultants, AI users completed 12.2% more tasks and worked 25.1% faster on tasks inside the AI capability range. On tasks outside that range, AI users were about 19 percentage points less likely to be correct.

Goldman Sachs said AI's contribution to U.S. GDP growth in 2025 was essentially zero. As of March 2026, Goldman found no meaningful relationship between AI and economy-wide productivity. The St. Louis Fed found workers reported AI saving 2.2% of all U.S. work hours in Q2 2026, up from 1.6% in Q3 2024.

What Skeptics Are Saying

MIT economist Daron Acemoglu, who won the Nobel Memorial Prize in Economic Sciences in 2024, estimates AI will deliver roughly 0.55% in total factor productivity gains over the next decade. That figure is far below projections embedded in current market valuations.

NBER data shows executives estimated AI's contribution to labor productivity growth over the past three years at just 0.2% per year, based on responses from nearly 9,000 executives who reported broad adoption but modest realized gains.

Despite all of this, 94% of companies plan to keep investing in AI even if it does not generate immediate returns, per BCG's AI Radar 2026. Companies expect to roughly double AI investment from 0.8% of revenue in 2025 to 1.7% in 2026. The spending continues. The verified proof does not yet match it.