In a coordinated move on January 27, 2026, the Midwest decided to bet the farm on the atom.
In a rare display of bipartisan urgency, Wisconsin’s Assembly passed AB 472, creating massive tax incentives for new nuclear construction, effectively declaring it the state’s “second priority” energy source behind efficiency. Simultaneously, Indiana’s Senate passed SB 258, a deregulatory sledgehammer that strips away state permitting requirements for reactors entirely.
The political narrative is clean and compelling: AI data centers are coming to the Rust Belt, they need 24/7 power, and wind turbines can’t provide it. Therefore, the region must unlock nuclear.
However, a fatal flaw exists in this logic. These bills solve a paperwork problem in a world constrained by physics.
Legislators can issue a permit today. They can offer a tax credit tomorrow. But statutes cannot reprint the laws of thermodynamics, nor can they conjure a supply chain that has been dormant for thirty years. The Midwest is legalizing a technology that, for all practical purposes, does not yet exist at the scale required.
This is the nuclear mirage: a legislative “green light” for a vehicle that has no engine and no fuel.
The Paperwork vs. The Physics
To understand why these bills are a trap, one must look at what they actually do versus what the industry actually needs.
Indiana SB 258 acts as a classic deregulatory play. It removes the Indiana Department of Environmental Management’s authority to hold public hearings on specific nuclear aspects, effectively fast-tracking the “License to Construct.” Wisconsin AB 472 offers a $10,000 per megawatt tax credit (over ten years), a sweetener designed to make the economics rival natural gas.
If this were 1970, these bills would be revolutionary. In 2026, they are largely irrelevant.
The bottleneck for nuclear energy in the United States is no longer regulatory per se. The Nuclear Regulatory Commission (NRC) has already approved designs like the AP1000 and NuScale’s SMR (Small Modular Reactor). The bottleneck is hardware.
The HALEU Void
The most immediate physical constraint is fuel. The next generation of reactors (specifically the SMRs that Microsoft and Google are courting for their data centers) do not run on the standard Low-Enriched Uranium (LEU) fuel rods that power today’s fleet. They run on HALEU (High-Assay Low-Enriched Uranium).
The math makes the Midwest’s ambition nearly impossible:
- Enrichment Level: Standard fuel is enriched to 3-5% U-235. HALEU is enriched to 19.75%.
- The Only Supplier: Until recently, the primary commercial supplier of HALEU was TENEX, a subsidiary of Russia’s state-owned Rosatom. That supply is now toxic due to sanctions.
- US Capacity: As of January 2026, the only licensed US company producing HALEU is Centrus Energy in Ohio.
- The Gap: Centrus is ramping up, but its capacity is currently measured in kilograms. The Department of Energy estimates the market needs 40 metric tons per year by 2030 to support a commercial SMR fleet.
The U.S. is currently meeting less than 3% of the projected demand. Passing a bill in Indiana to “fast track” a reactor that requires fuel the country cannot produce is akin to rezoning land for a gas station in a world without gasoline. It relies on a “Field of Dreams” energy policy: If you permit it, they will come. But the industry cannot come if it cannot fuel up.
More critically, the transport infrastructure for HALEU is missing. Standard uranium casks (Type A) cannot legally carry 19.75% enriched fuel due to criticality risks. The industry needs Type B casks, heavily shielded containers that require their own separate, years-long NRC certification process. As of Q1 2026, the fleet of certified Type B casks available for commercial HALEU transport is effectively zero.
The Time Dilution: 60 Months is a Lie
Assuming utilities stick to traditional large-scale reactors like the Westinghouse AP1000 (the gigawatt-scale beasts that don’t need HALEU), the situation remains grim. These are proven technology, certainly. But the construction timelines tell a different story.
The industry standard pitch (repeated by Cameco’s COO in early January) involves a 60-month (5-year) construction timeline. This is the “nth-of-a-kind” theoretical speed.
By contrast, the reality of Vogtle Units 3 & 4 in Georgia, the only new AP1000s built in the US in decades, is stark:
- Construction Start: 2013
- Unit 3 Online: July 2023
- Unit 4 Online: April 2024
- Total Time: ~11 years.
- Total Cost: ~\35 Billion (more than double the estimate).
If a Wisconsin utility breaks ground on an AP1000 in January 2026, and “learns” from Vogtle to shave 30% off the timeline, that reactor comes online in 2033.
The Lost Decade
This timeline fails the specific customer these bills are trying to attract: AI Hyperscalers.
Data centers are capital-intensive projects with a 2-3 year build cycle. If Microsoft breaks ground on a $5 billion cluster in Madison in 2026, they need power by 2028, not 2033. They cannot wait for a nuclear reactor.
This creates a perverse incentive. By passing these bills, state legislators get to claim they are “solving” the energy crisis with “clean, baseload nuclear.” When the data centers arrive in 2028 and the nuclear plants are still just holes in the ground, the utilities will have no choice but to fire up (or extend the life of) coal and natural gas peaker plants to meet the load.
The nuclear bills, ironically, provide the political cover to extend the life of fossil fuels.
The Capital Risk: Who Pays?
The most dangerous part of the “Nuclear Mirage” is the financial structure.
Wisconsin’s AB 472 provides tax credits, but it doesn’t cap ratepayer liability. In the Vogtle saga, Georgia ratepayers saw their bills climb significantly to cover the cost overruns. Nuclear construction is notorious for its “negative learning curve” in the West; costs tend to rise, not fall, as projects get complex.
For nuclear, is historically near 100%. By encouraging utilities to pursue these mega-projects, Midwest states are inviting a repeat of the V.C. Summer disaster in South Carolina, where an abandoned $9 billion project left ratepayers on the hook for billions without generating a single watt of power.
The Investor’s View
Wall Street recognizes this risk. Private capital has largely shunned new nuclear without 100% government backstops. The recent DOE announcement of an $80 billion push for AP1000s attempts to de-risk this, but even federal loans don’t print labor or pour concrete.
The smartest money in the room (the hyperscalers themselves) is hedging. Microsoft effectively bought Three Mile Island Unit 1 (the Constellation deal) not because they want to build new nuclear, but because they want the existing nuclear that actually works. They are buying the brownfield assets because the greenfield promises are vaporware.
The Labor and Waste Stalemate
Even if the fuel and money appear, two final barriers remain: people and waste.
The “Nuclear Renaissance” faces a critical shortage of qualified labor. Welding nuclear-grade piping requires specific “N-Stamp” certifications. The generation of welders that built the current US fleet in the 1970s and 80s has retired. Rebuilding that workforce takes almost as long as building the reactors themselves.
Furthermore, neither Indiana SB 258 nor Wisconsin AB 472 addresses the radioactive elephant in the room: waste storage. The US still has no central repository at Yucca Mountain. Every new reactor becomes its own de facto nuclear waste dump, storing spent fuel in dry casks on-site indefinitely. Local opposition to “forever waste” has historically derailed projects faster than any funding shortfall.
The Verdict: A Deregulatory Placebo
Wisconsin AB 472 and Indiana SB 258 are not bad bills in a vacuum. Removing bureaucratic red tape is generally positive. If safe, modern nuclear can be built, it should be legal to do so.
But framing these bills as a “solution” to the 2026 energy crunch is malpractice. They are a Call Option on the 2030s.
For the next five to ten years, the Midwest’s energy reality will be defined by what can actually be deployed: wind, solar, batteries, and natural gas. The nuclear bills are a mirage that allows politicians to point at a shimmering future while the grid struggles in the cold light of the present.
The lights will stay on in the Midwest, but they won’t be powered by these phantom reactors. They’ll be powered by whatever could effectively be bolted to the ground in 24 months. And right now, that isn’t nuclear.
🦋 Discussion on Bluesky
Discuss on Bluesky