Big Tech has finally stopped pretending: the AI boom isn’t just about compute, GPUs, or “the future of work.” It’s about electricity, staggering, nation‑reshaping amounts of it and the companies that want to secure their own energy futures are now embedding themselves directly into the power grid.
The latest example comes from Microsoft, which this week announced a partnership with the Midcontinent Independent System Operator (MISO), the grid operator serving 42 million people across 15 U.S. states and Manitoba. The deal, framed as a benevolent modernization effort, will use Microsoft Azure and Microsoft Foundry AI to build a “unified data platform” for grid forecasting, analytics, and resilience.
If that sounds familiar, it’s because it is. Google inked a similar arrangement with PJM Interconnection, the country’s largest grid operator, last year, using AI to accelerate the onboarding of new electricity supplies to meet the industry’s ballooning power appetite.
This isn’t a trend. It’s a takeover.
Microsoft’s messaging is predictably polished. MISO’s Chief Information and Digital Officer Nirav Shah gushes that partnering with Microsoft will “harness the full power of advanced analytics, AI and cloud platforms to improve forecasting, enhance decision-making and build resilience into our operations”. He adds that such acceleration is “critical because of the increasing diversity of energy mix, electrification, rising demand and the growth of data centers”.
Microsoft’s own Darryl Willis, corporate VP for energy and resources, frames the partnership as a public service: “By bringing advanced cloud and AI capabilities to grid operations, we’re building a future-ready, more resilient and sustainable grid that can anticipate challenges, optimize performance, and deliver reliable power as electrification and demand grow”.
It’s the kind of language that makes you feel like Microsoft is doing the Midwest a favor, swooping in with cloud magic to save the grid from collapse.
But let’s be honest: this isn’t philanthropy. It’s self-preservation.
AI’s Energy Hunger Is the Real Story
The Reuters report on the deal doesn’t even bother with subtlety: this partnership is “the latest instance of Big Tech turning to collaboration to help ensure the massive amounts of electricity needed for artificial intelligence are available”.
Translation: AI data centers are consuming so much power that tech companies are now integrating themselves into the grid to guarantee supply.
Over the last two years, the industry has been “locking into long-term supply agreements” and offering AI tools to utilities to “maximize supplies” as U.S. power demand hits record highs.
This is not modernization. This is resource capture.
States Are Handing Over the Keys
The most alarming part isn’t that Microsoft wants to secure energy for its AI empire. It’s that states and regional grid operators are rolling out the red carpet.
When a grid operator serving 42 million people partners with a trillion‑dollar corporation whose business model depends on exponential energy growth, that’s not collaboration, it’s collusion.
These partnerships set a dangerous precedent, quietly shifting public‑interest energy planning into the hands of the very corporations driving unprecedented demand. When Microsoft’s AI tools start shaping forecasting, planning, and operational decisions, the company effectively becomes a shadow energy regulator, influencing which resources get prioritized, how quickly transmission lines are built, and where scarce capacity ultimately flows.
And once a grid operator’s core systems are rebuilt on Azure, the region is effectively locked into a private cloud platform whose incentives skew toward corporate growth, not public resilience. From there, the trajectory becomes predictable: energy planning bends toward the needs of hyperscale data centers rather than the communities they sit beside, because the loudest voice in the room is always the one demanding gigawatts for AI expansion.
Meanwhile, the public ends up subsidizing the infrastructure upgrades required to keep these data centers humming. New transmission lines, substations, and reliability improvements don’t materialize out of goodwill, they show up on ratepayer bills, not Microsoft’s balance sheet.
And if this all feels unprecedented, it isn’t, it’s simply the Midwest catching up to a pattern Big Tech has already begun carving into the nation’s energy infrastructure. Microsoft isn’t the first company to embed itself upstream in the grid, and it certainly won’t be the last. The industry has been testing the boundaries of how far it can push public‑private “partnerships” before regulators realize they’re ceding control of essential infrastructure to corporations whose primary goal is securing power for AI expansion. And the clearest early warning sign didn’t come from the Midwest at all, it came from the East Coast.
Google’s partnership with PJM Interconnection, the largest grid operator in the United States was the first unmistakable signal that Big Tech wasn’t content to merely consume electricity. It wanted influence over how the grid itself evolves. PJM oversees power for 65 million people, and Google’s AI tools are now woven into its planning process to “speed up the process of connecting new electricity supplies.” On paper, that sounds like a win for modernization. In practice, it’s a polite way of saying the grid needs to accelerate the onboarding of generation because hyperscale data centers are devouring power faster than utilities can build it.
“New electricity supplies” isn’t some abstract, civic‑minded goal. It’s code for more capacity to feed AI clusters, the same clusters that have already strained local grids, delayed renewable projects, and forced utilities to rethink decades‑old planning assumptions. Google didn’t partner with PJM to help grandma keep her lights on during a heat wave. It partnered with PJM to ensure its AI roadmap isn’t throttled by the physical limits of the grid.
Seen in that light, the Microsoft–MISO deal isn’t a bold new direction. It’s the next domino in a chain reaction Big Tech set off the moment AI’s energy appetite became too large to hide behind sustainability reports and glossy keynote slides. The companies that need the most power are now embedding themselves inside the institutions that decide how power gets built, allocated, and priced.
And once you see the pattern, Google embedding itself in PJM, Microsoft embedding itself in MISO, it becomes impossible to pretend these are isolated modernization projects. They’re coordinated moves in a broader shift where the companies consuming the most power are positioning themselves to influence how that power gets produced, allocated, and priced. Each new partnership normalizes the idea that Big Tech should sit inside the control room of public infrastructure, shaping decisions that used to be made by regulators, planners, and communities. And that brings us to the uncomfortable truth policymakers keep dodging.
States should not be entering into deep operational partnerships with the very companies driving unprecedented strain on the grid. It’s the regulatory equivalent of letting a mining company help rewrite environmental rules, the conflict of interest isn’t subtle, it’s structural. The grid is a public good. AI is a private profit engine. When the two become intertwined, the public loses leverage, transparency, and control.
Microsoft and Google aren’t modernizing the grid out of civic duty or regional loyalty. They’re modernizing it so it can feed the insatiable energy demands of AI systems that generate billions in revenue and require more electricity than entire cities. And unless regulators wake up to the scale of what’s happening, the future of American energy won’t be shaped by public planning or democratic oversight. It will be shaped by corporate necessity, and by the companies that got there first.
