AI Capex Is Booming. The Broad Investment Boom Isn't.

AI spending is lifting headlines, but net investment still looks weak

Investors are buying a broad prosperity story, but the macro data still point to a narrow AI boom. AI spending is real: computer equipment investment rose from 0.5% to 1.3% of GDP in under three years. Yet total net investment is near its lowest share of GDP outside the 2009 crash. That matters because spending can keep existing equipment running without clearly expanding the economy's future income base.

That selective surge is already biasing growth headlines. the technology is already affecting gross domestic product (GDP) numbers through its associated investment boom. And the spending is concentrated, not diffuse. AI capex is meaningful relative to the $30 trillion U.S. economy, and one estimate suggests official data may understate the outlay by the companies leading the data-center buildout. That can make a narrow spending surge look like broad prosperity.

Markets are amplifying that impression. the strength in the /ES has been a huge pitcher of ice water into the face of the bears, while the /NQ... has a well-formed symmetric triangle which is being challenged at this very moment. The practical issue is simple: if investors keep pricing AI-led momentum as if it were economy-wide wealth creation, the next macro data checks become more important.

The spending is real, but the economic payoff is still unproven

Real spending, uncertain value added

Investment in computer equipment has exploded from 0.5% of GDP to 1.3% in under three years, and that surge is already showing up in macro data. Broader estimates also tie a large share of recent growth to AI infrastructure, including one read that AI infrastructure investment accounted for 92% of US GDP growth in the first half of 2025 and a St.

Louis Fed measure that attributed a much more modest 39% of total GDP growth to AI-related investment in the third quarter of 2025. The boom is not imaginary.

The harder question is what value this spending is creating. The current evidence framework for major technologies suggests capabilities and cost declines precede broad firm adoption and investment, which in turn precede measurable aggregate productivity gains and labor market outcomes. For now, the AI buildout still looks more like the investment phase than a confirmed productivity boom.

Concentration can look like a wealth effect

That concentration is investable, but it is not the same as an economy compounding wealth across sectors. AI-related spending is meaningful relative to the $30 trillion U.S. economy, and existing methodologies are (probably) not fully capturing the investment being done by the five companies responsible for the bulk of the data center buildout. That helps explain why the market can look strong even while the broader capital story remains narrow.

If that spending proves profitable and spreads beyond a handful of giants, the case for a broader boom will strengthen. If not, today's investment surge may look more like a costly buildout than a durable wealth effect.

Adoption is rising faster than captured economics

Usage is broad. Monetization is less clear.

The capex story is already established. What investors may be overreaching on is the leap from adoption headlines to durable profit streams. By August 2025, generative AI tools were used by 55% of people and 37% of workers in the U.S.. That is fast adoption, but usage alone does not prove that companies are capturing lasting economic value.

It is easy to confuse widespread use with strong monetization. Consumer and worker adoption can improve workflows without immediately showing up in stronger margins, cleaner profit concentration, or broad-based earnings power.

The missing bridge is productivity

The evidence framework matters here. The AI buildout literature explicitly tracks capabilities and costs; firm adoption and investment; and productivity and labor as a sequence. Adoption is only part of the path. The more important valuation question is whether AI eventually produces measurable aggregate productivity gains and labor market outcomes.

That is the clearest gap between the current narrative and the current evidence. Investors do not need to deny that AI spending and adoption are real. They need to distinguish between early uptake and proven economic returns.

What would confirm the boom - and what would challenge it

Signals that support the narrow-boom view

• Net investment stays weak. If total net investment is near its lowest share of GDP outside the 2009 crash, the economy still looks more like it is maintaining capital than broadly expanding it.
• Markets stay narrow. the strength in the /ES has been a huge pitcher of ice water into the face of the bears, but the story remains most constructive where AI and mega-cap tech leadership dominate.
• AI spending remains a large driver of growth. Estimates still tie a very large share of recent expansion to AI infrastructure investment.

Signals that would weaken the narrow-boom view

• Net investment broadens. A more durable rise in total net investment would suggest the economy is building beyond a narrow AI corridor.
• Productivity shows up. The clearest confirmation would be measurable aggregate productivity gains and labor market outcomes.
• Monetization spreads. Widespread use of generative AI tools were used by 55% of people and 37% of workers in the U.S. is only the start; the next step is clearer evidence that firms are turning that adoption into sustained economic gains.