Citi Identified Power Bottlenecks for AI Spending

Power generation capacity may limit infrastructure growth despite billions in projected capital investments.

Updated on Oct. 5, 2026 in Artificial Intelligence

Bold flat-color editorial illustration of steel high-voltage towers in a repeating rhythmic formation, representing industrial infrastructure constraints. — Citi strategist Stuart Kaiser warns that electrical power generation constraints pose a persistent bottleneck for artificial intelligence infrastructure despite projected capital spending. AI Illustration. Upload story photo >

Live Poll

Do you believe power infrastructure limits will significantly slow the long-term adoption of AI technology?

Citi strategist Stuart Kaiser warns that power generation constraints remain a significant hurdle for artificial intelligence growth. These infrastructure limitations persist even as the firm projects nearly $4 trillion in total capital spending by 2030.

Why it matters

Rising costs and utility supply bottlenecks create a divergence between AI infrastructure demand and energy availability. This environment shifts focus toward how power constraints impact the broader market's ability to realize earnings growth.

Investment flows show a 7.07% year-to-date rise in the XLV healthcare ETF and a 4% increase in the POWR infrastructure ETF, while the XLU utilities ETF has declined over 7% year-to-date.

The players

Citi

A global financial services firm providing market research and capital expenditure analysis for the technology sector.

Stuart Kaiser

A strategist at Citi who analyzes how capital expenditures and utility constraints influence market performance.

The details

Rising yields and election-related affordability concerns are currently exerting pressure on utility providers in the United States. These market forces, combined with physical limitations on available megawatts, create a bottleneck that slows the immediate scaling of data center hardware. Meanwhile, the integration of artificial intelligence in healthcare drug development continues to present distinct growth opportunities that operate independently of these power supply constraints.

September 2026: Citi issued caution regarding rising utility and power costs.
Next 12 to 24 months: AI-related capital spending remains committed.
2027: Estimated $1 trillion in AI-related capital spending.
2030: Estimated nearly $4 trillion in total AI capital spending.

The Tech Race

This analysis positions power generation as the primary gating factor in the multi-trillion-dollar race to build massive AI infrastructure. It departs from previous optimistic growth models by grounding performance expectations in the current capacity constraints of the U.S. power grid.

Investors should monitor utility sector volatility and infrastructure fund performance as power generation constraints persist over the next 12 to 24 months. These market dynamics suggest that growth-oriented portfolios may face performance disparities tied directly to local grid capacity.

The takeaway

The trajectory of AI expansion is currently defined by a conflict between multi-trillion-dollar capital investment and a restricted energy supply. Watch the quarterly performance of infrastructure-focused ETFs to see if utility supply catches up to the 2027 spending targets.