President Trump is about to sit down with the CEOs of Google, Meta, OpenAI, Microsoft, Oracle, and Amazon for a conversation that could reshape how AI gets built in America.
The topic? Electricity bills. Specifically, how to stop Big Tech’s AI spending spree from jacking up your power costs.
Trump’s “Ratepayer Protection Pledge” sounds boring, but the implications are massive. These companies are burning through electricity faster than small countries, and somebody has to pay for the grid upgrades. Until now, that somebody was regular consumers.
The pledge forces tech giants to bring their own power — either by building new plants, buying existing capacity, or paying for grid upgrades themselves. No more socializing the costs while privatizing the AI profits.
For investors, this is huge. It means data center companies and energy infrastructure plays just got a permanent revenue tailwind. Every hyperscaler will need dedicated power deals, custom grid connections, and industrial-scale backup systems.
It also means the AI buildout just got more expensive for Big Tech. Meta and Google can’t just plug into the local grid anymore — they need to invest billions in power infrastructure upfront. That’s a margin headwind for the hyperscalers and a revenue goldmine for energy companies.
The timing isn’t accidental either. With midterms approaching, energy costs are a political third rail. Trump is forcing tech CEOs to choose: keep building AI data centers at this pace but pay the full freight, or slow down.
Smart money is already rotating into the picks-and-shovels plays: Constellation Energy for nuclear power, Vertiv for cooling systems, GE Vernova for grid infrastructure. The companies that make AI physically possible just became a lot more valuable.
Wall Street hasn’t fully priced this shift yet. When hyperscalers start signing 20-year power purchase agreements and billion-dollar infrastructure deals, these energy stocks will look ridiculously cheap in hindsight.