Three years ago, Ford, GM, and Honda were pouring billions into battery plants built for one purpose: keeping up with EV demand that industry planners assumed would keep climbing in a straight line. It didn’t. Now those same buildings are getting rewired for a different customer entirely — the data centers running the country’s AI boom.
You don’t retool a billion-dollar factory on a whim. What’s happening across the battery supply chain right now is a direct read on how badly automakers overbuilt for an EV ramp that federal tax-credit cuts and softer-than-projected consumer demand never let materialize — and how fast that capacity is finding a second life powering server farms instead of SUVs.
1. Honda: from canceled EVs to data-center batteries in three months
Honda’s clearest pivot is its Ohio joint venture with LG Energy Solution, built to supply cells for a slate of American-market EVs. According to TechCrunch’s reporting, Honda canceled three of those EV programs around April 2026 after federal EV tax credits were eliminated the previous September, then redirected the Ohio plant toward stationary energy-storage batteries rather than unwind the joint venture. Production for data centers and grid storage began in early July 2026. TechCrunch notes the stationary storage market Honda is chasing has grown 32% year over year and carries better margins than the EV batteries the plant was originally built to produce.
2. GM: an LFP line built for storage, not cars
GM’s Ultium Cells joint venture with LG Energy Solution is running a lithium iron phosphate (LFP) line at its Spring Hill, Tennessee plant explicitly for stationary storage rather than vehicle packs, a shift GM Authority reports gives GM “another source of demand at a time when EV adoption has cooled more quickly than originally anticipated.” GM has also partnered with battery recycler Redwood Materials, whose Nevada microgrid operation is already using GM-made cells to help power AI infrastructure. LFP chemistry suits stationary storage well: the weight and packaging constraints that matter in a car are far less important bolted to the side of a data center.
3. Ford: a battery plant’s second act as a grid asset
Ford is redirecting an EV battery plant in Glendale, Kentucky toward grid-scale energy storage, signing renewable power developer EDF as its first customer and investing $2 billion to scale the operation toward a target of at least 20 GWh of annual output by late 2027. Ford’s move sits inside a wider industry trend: BloombergNEF has tracked at least 11 battery plants, eight of them in the US, being retooled from EV production toward energy storage.
The overcapacity math
None of this is happening because automakers suddenly discovered a passion for server farms. It’s happening because the battery capacity is already built and paid for, and EV sales growth in the US has come in well below the projections that justified building it. Redirecting that output toward grid storage and AI-adjacent power infrastructure — a sector with genuinely insatiable, immediate demand — is a way to put idle or underused capacity to work without writing off the investment entirely. Stationary storage doesn’t require winning back a skeptical car buyer; it just requires a utility or a data-center operator willing to sign a supply contract, which right now is not the hard part of the AI buildout.
The community pushback that’s actually slowing things down
The hard part is what happens after the batteries ship: getting the data centers themselves built. Data Center Watch, an industry-tracking group, found that at least 75 data center projects worth roughly $130 billion were blocked or delayed in the first quarter of 2026 alone, driven by local opposition in 49 states. Fourteen states saw statewide moratorium proposals introduced, and more than 300 state-level data center bills were filed in just the first six weeks of the year. Maine came within a single House vote of a statewide ban. The group’s data shows organized opposition groups more than doubled compared with the end of 2025, with petition-signature volume in the first quarter nearly matching all of the second half of last year combined.
The objections tracking that opposition center on power draw, water consumption for cooling, land use, and the fear that ordinary ratepayers will end up subsidizing the electricity a data center consumes. Those are the same grid-capacity concerns that make automaker-supplied battery storage valuable in the first place — a data center that can pair with on-site or nearby battery storage has an easier time making its case to a skeptical planning board than one asking a strained local grid to simply absorb the load.
That’s the loop these battery plants have landed in. Automakers built capacity for EVs that didn’t sell as fast as forecast, found a ready buyer in grid storage and AI infrastructure, and now that infrastructure is running into the same kind of local resistance that has slowed wind farms, transmission lines, and warehouses before it. The batteries aren’t the bottleneck anymore. The permits are.