Key Highlights
- A basket of AI infrastructure stocks, dubbed "AI Enablers," has outperformed the MSCI World index by 50% over the past three years.
- US Data Center developers face a projected 40-gigawatt grid access shortfall through 2028, roughly half the power capacity needed.
- Power transformer delivery times in the US now average more than two years, up from a pre-pandemic norm of 12 to 16 weeks.
- AI enablers in the US captured 65% of Capital raised in seed to late-stage venture funding rounds in the first half of 2026.
Grid Constraints Drive Investor Interest in Infrastructure
Investors are increasingly favoring companies that provide infrastructure supporting AI development, as electricity Supply constraints emerge as a key bottleneck to growth. A basket of AI infrastructure stocks spanning utilities, information technology, and hyperscalers has outperformed the MSCI World Index by 50% over the past three years. AI enablers in the US captured 65% of capital raised in seed to late-stage venture funding rounds in the first half of 2026, reflecting investor confidence in continued Demand growth.
Token usage, the units of data processed by AI models, rose tenfold between January 2025 and April 2026, according to Fintech company Ramp.
Grid Access Shortfall Poses a Structural Challenge
Analysis projects a roughly 40-gigawatt grid access shortfall for US data center developers through 2028, representing about half of the power capacity needed to meet demand. Power transformer delivery times in the US now average more than two years, compared with a pre-pandemic norm of 12 to 16 weeks, and some utilities are quoting interconnection timelines of a decade or more in certain regions. Political pushback in some areas over rising power bills and concerns about air quality and water availability has added further delays to data center construction.
Fuel Cells and Battery Storage Seen as Partial Solutions
Providers of fuel cells, which convert energy sources into electricity and can often be installed faster than new grid connections, are seen as one potential solution, since they can operate off-grid with limited water use and fewer emissions. Battery energy storage systems are also emerging as a system-level necessity, allowing data centers to manage peak power demand, defer costly infrastructure investment, and participate in grid stabilization efforts. Analysts note that AI's shift from model Training toward real-time inference and agent-based tasks is creating more volatile power demand patterns that existing grids were not designed to accommodate.





_11_20_2026_12_38_15_908731.jpg)
Please wait processing your request...