The AI Infrastructure Play Hiding in Plain Sight That Could Be 2026's Biggest Surprise

NextEra Energy is positioned as an AI infrastructure beneficiary, with management citing 60% electricity demand growth from 2025 to 2045 and a dividend yield of 2.8% versus a 2.5% utility average. The company is also pursuing Dominion Energy, a deal that would expand regulated operations into four states from one and improve exposure to Virginia data center demand. The article is broadly supportive of NextEra’s long-term growth and income profile, though the Dominion acquisition still faces a lengthy regulatory approval process.

Analysis

The market is still treating AI power demand as a generic utility bull case, but the real edge is in grid-access scarcity, not megawatt growth itself. NextEra’s advantage is that it can monetize both regulated load growth and merchant/contract capacity, which means it is better positioned than pure-play utilities to capture the first wave of data-center interconnect demand before new transmission and generation can be built. If hyperscalers keep signing multi-year power deals, the winner is less about the cheapest kWh and more about who can deliver firm capacity on a credible timetable.

The second-order effect is that this thesis likely compresses returns for smaller renewable developers and merchant generators without balance-sheet scale. AI load is supportive for the sector, but it will also expose which utilities have weak execution on transmission, permitting, and long-duration capital allocation. Dominion’s asset footprint would matter less for immediate earnings than for option value: Virginia is a strategic node because data-center clustering tends to reinforce itself, and once a utility is embedded in that ecosystem it can win incremental load with lower customer acquisition cost.

The biggest miss in consensus is duration. Investors are extrapolating a multi-year demand ramp, but the stock can still re-rate before cash flows fully show up if the market believes rate base growth and contracted renewables are becoming structurally more valuable. The main reversal risks are regulatory delay, rate-base dilution from a large acquisition, or a pause in AI capex that pushes out utility load expectations by 12-24 months. In that scenario, the market could quickly stop paying up for the AI narrative and revert to pure utility valuation multiples.

This is attractive as a quality-growth defensive, but not as a clean catalyst trade. The setup is better for owning NEE into volatility than for chasing a headline breakout; if the merger process drags, the equity should still collect carry through the dividend while investors wait for optionality to accrue. Relative value is strongest against lower-quality utilities and against infrastructure names that lack contracted demand exposure.

Key Decisions for Investors

• Long NEE on a 6-12 month horizon; use pullbacks toward utility-sector underperformance as entry points. Risk/reward is favorable because downside is buffered by dividend carry, while upside comes from multiple expansion if AI load data continues to inflect.
• Pair trade: long NEE / short a lower-growth regulated utility ETF or a single-state utility with limited data-center exposure. The trade expresses dispersion between scaled grid owners and utilities that lack the balance-sheet or geography to capture AI load.
• Avoid paying up for D ahead of merger clarity; if using it, prefer a market-neutral merger-spread structure rather than directional long exposure. The key risk is regulatory slippage, which can keep capital trapped for quarters without matching earnings acceleration.
• Sell downside protection on NEE only if implied vol stays elevated and the stock remains below a premium utility multiple. The dividend cushions carry, and the market may continue to underprice the option value of AI-linked load growth.
• If AI capex sentiment rolls over, rotate out of pure-renewables developers and into NEE. NEE has the best mix of contracted cash flow and growth optionality, making it the least fragile way to own the power-demand theme.