Key Takeaways

  • Rapid AI investment is consuming an unusually large share of US capital spending
  • Analysts warn that uneven productivity gains and labor impacts are slowing broader economic growth
  • Heavy spending on foundation models is increasing concentration risk across key digital markets

A growing share of US capital spending is being pulled into artificial intelligence projects, particularly large-scale foundation model development. The trend accelerated through 2024 and 2025, and by mid-2026, several traditional sectors are feeling the strain from reduced investment availability.

AI-related capital spending in the United States exceeded $250 billion in 2024. This investment accounted for about half of reported GDP growth, or roughly 1.2%, in the first half of 2025. When spending is narrowly focused, it creates ripple effects in credit markets, hiring cycles, and corporate R&D budgets. Some smaller manufacturers and service providers have slowed expansion while watching input costs rise.

AI has not yet delivered the broad productivity lift many economists once predicted. According to Holistic Data Solutions, which summarizes work by PwC, global AI could add up to $15.7 trillion to GDP by 2030. The gap between potential and current results, however, is wide. Many anticipated benefits rely on long-run organizational shifts that are slow and require complementary investment in data infrastructure, workforce skills, and process redesign. Until those operational changes take root, much of the capital poured into AI remains speculative.

Labor market effects add another layer of complexity. Analysts at the International Monetary Fund, summarized by MIT Sloan, estimate that AI will affect about 40% of jobs globally. Advanced economies are expected to face higher automation pressure, which may amplify inequality if wage gains concentrate at the top. The current wave of automation risks simultaneously depressing wages for entry-level roles while raising compensation for a narrow set of highly skilled technical positions. The distributional picture remains complex as companies navigate the right mix of augmentation and automation.

In parallel, roughly 300 million full-time jobs worldwide are exposed to generative AI automation, according to a Goldman Sachs analysis cited by IEDC. Exposure does not automatically mean displacement, but it signals the scale of potential disruption. Workforce transition programs take time to design and fund, and businesses that rely on stable clerical or operational roles face difficult planning choices.

Technology cycles often create pockets of overinvestment, similar to the internet boom of the late 1990s. What differentiates the current environment is the extreme level of capital required to train frontier-scale models. Companies like OpenAI, Anthropic, and Google operate with compute budgets that smaller competitors cannot realistically match. This reinforces scale advantages and leads to what some economists are calling an AI infrastructure bubble. The Washington Center for Equitable Growth flagged this dynamic in 2025 when examining how much of the country's economic growth was tied exclusively to AI-related capital formation.

Because few organizations can afford that level of spend, the competitive landscape is tilting toward a handful of infrastructure providers. This narrows innovation channels, leaving customers dependent on a small group of vendors. Concentration risk rises when the underlying technology and supply chains converge so tightly. Executives across multiple industries are reconsidering how they evaluate vendor diversification as the largest market players continue to accelerate hiring and hardware procurement.

The regulatory and standards environment is attempting to keep pace with these market concentrations. The NIST AI Risk Management Framework offers guidance on identifying and reducing model-level risks, and the OECD AI Principles encourage broad-based economic benefits rather than extractive use patterns. Organizations are adopting these frameworks to structure internal governance and align external reporting. While they do not provide a full economic blueprint, the frameworks push organizations to evaluate the second-order effects of deployment.

The current wave of AI investment carries a steep opportunity cost. When financing and internal budgets flow disproportionately into AI, other corporate initiatives lose ground. Renewable energy developers and mid-sized biotech firms have reported project delays tied to restricted capital availability. Enterprise software vendors outside of the AI core have also seen customers push contract renewals into later quarters to reallocate funds for model experimentation.

The Federal Reserve's April 2026 monitoring note highlighted that AI adoption varies significantly by sector. Certain industries are realizing returns from targeted automation, while others remain stuck in pilot phases. This uneven distribution of benefits contributes to the economic strain identified in mid-2026, as the rapid technological shift fails to uniformly reward all market participants. For business leaders, the operational challenge involves deploying AI without undermining core strategic priorities.

Market pressures are likely to intensify as companies confront the gap between initial expectations and realized productivity gains. Although overall investment optimism remains high, capital markets are scrutinizing whether the next phase of AI deployment will spread economic benefits broadly or keep funding locked inside a small set of upstream technology providers. That tension will directly shape corporate investment patterns across the remainder of 2026 and beyond.