Right to Compute: The Legal Test Case Data Center Buyers Should Watch

On November 4, 2026, a Missoula, Montana AI company called CubCloud AI filed a lawsuit most of the data center industry hasn’t noticed yet, but should. The company is suing its own city over a zoning ordinance that blocks new data center development, arguing the rule violates Montana’s “Right to Compute Act” — the first state law in the country that treats access to computational infrastructure as a protected right. It is also, as of this month, the first time that law has ever been tested in court. The timing is not a coincidence. The same week the lawsuit landed, trackers counted 583 active local and state restrictions on data center development across 46 states, up from roughly $130 billion in projects delayed or blocked in just the first quarter of the year. Buyers who treat community and regulatory opposition as a single, predictable line item are behind the curve: the fight has split into two tracks, and the second one — state laws designed to override local control — is now in front of a judge for the first time.

The restriction count keeps climbing, and it’s not slowing down

As of November 4, 2026, trackers following the issue counted 583 active restrictions on data center development spread across 46 states, a figure researchers describe as “a floor rather than a precise census” because moratoriums are routinely extended, quietly lapse, or get replaced with something stricter before anyone updates a map. More than 350 of those 583 are local actions — city councils, county boards, and townships — rather than state-level rules, and the mix keeps shifting toward permanent bans rather than temporary pauses: St. Charles, Missouri; Walton County, Florida; Plainfield, New Jersey; Woodford County, Kentucky; and Cumberland, Rhode Island have all enacted permanent restrictions since May. Temporary moratoriums are getting longer, too. Cary, North Carolina passed an 18-month pause running through February 2028. Indianapolis extended its hold through the end of 2027. Montgomery County, Maryland adopted an 18-month freeze in July. These aren’t symbolic gestures by small towns; they’re multi-year windows in which a buyer simply cannot permit new load, no matter how clean the power contract or how fast the interconnection study moves.

State legislatures are adding a second layer on top of local action. Michigan has six statewide data center bills pending. Vermont, Minnesota, and Maryland have statewide bills in committee. Twenty-four states now have pending legislation of some kind, and four federal bills are moving through Congress, including the Ratepayer Protection Act, which passed the House 417–3 in September before stalling in the Senate. None of this reads as a market correcting itself after a few loud town halls. It reads as a structural shift in how much friction a new site carries before a shovel goes in the ground, and that friction is compounding rather than fading as 2026 closes out.

Enter “right to compute”: the legal counter-movement

Montana passed the first statute of its kind in April 2026. The Right to Compute Act (SB 212) says government limits on owning or using “computational resources” — defined broadly enough to cover servers, data processing, and storage infrastructure — must be “demonstrably necessary and narrowly tailored to fulfill a compelling government interest.” That’s a high constitutional bar, deliberately borrowed from strict-scrutiny language used elsewhere in civil-rights law, and it was not written in a vacuum: the bill closely tracks model legislation from the American Legislative Exchange Council, which means it can be copied into other statehouses with minimal rewriting. New Hampshire, Ohio, and South Dakota are now considering versions of it; a similar push in Idaho stalled in committee.

Supporters frame these laws as a check on local governments overreacting to AI anxiety with blunt zoning tools. Critics, including several state legislators who voted against Montana’s bill, argue the practical effect is to hand large infrastructure developers a legal override button against the same city councils and county boards that are supposed to represent the residents living next to a 300-megawatt campus. Both readings can be true at once, and that tension is exactly why the Missoula case matters: it’s the place where the theory meets a real zoning map for the first time.

The first test case: Missoula v. CubCloud AI

CubCloud AI is a small company, 18 employees, founded roughly a year ago, running an existing facility near Missoula’s airport that complied with the city’s old zoning rules because its computing equipment occupied less than half the building’s floor area. In October, the Missoula City Council closed that loophole, rewriting the definition of a data center to regulate facilities by how their equipment is used rather than by how much floor space it occupies — a change that would block the stand-alone building CubCloud wanted to add next. The company, represented by the Goldwater Institute’s Big Sky Legal Project, is asking a court to declare the ordinance unlawful under the Right to Compute Act and to issue an injunction stopping the city from enforcing it. The complaint argues the city has no “compelling governmental interest” in the restriction and that, even if one exists, a blanket rule isn’t narrowly tailored to it, noting the facility generates no unusual noise, odors, or nuisance complaints.

Missoula’s city attorney, Ryan Sudbury, confirmed the city plans to defend the ordinance once formally served, and made a point worth sitting with: the Right to Compute Act “has not been tested in court.” That’s the whole story in one sentence. Every buyer currently counting on a right-to-compute statute as a backstop against a hostile local council is relying on a legal theory with zero case law behind it. It may hold up. It may not. Nobody actually knows yet, and won’t for months.

Memphis shows how fragile these votes really are

If the Missoula case shows the legal fight in its early innings, Memphis shows how little margin exists on the political side even without a competing statute in play. On October 20, the City Council deadlocked 6–6 on an 18-month moratorium covering new data centers and expansions, including the facilities tied to Elon Musk’s xAI and its Colossus supercomputer operation. The measure failed only because Councilwoman Yolanda Cooper-Sutton was absent; a single vote either way would have flipped the outcome. Nothing replaced it at the council level, though Mayor Paul Young separately ordered a 90-day review to recommend rules for large-scale facilities, running in parallel with continued council deliberation. Sponsor JB Smiley Jr. was explicit about the intent: “We’re not saying no to data centers forever; we’re saying slow it down so that we can make the necessary adjustments.” Council member Jerri Green called the failed vote a loss for residents who had been pushing for it. Opposition in Memphis centers on air quality and gas-turbine emissions tied to xAI’s operations, concerns serious enough that the NAACP filed a federal Clean Air Act lawsuit over turbines at a related Southaven, Mississippi site — a case the Justice Department has since urged a court to dismiss, with the matter still active. The lesson for buyers isn’t that Memphis is safe; it’s that “the moratorium failed” and “the moratorium passed” are often separated by one absent vote, one amendment, or one election cycle, which makes political handicapping in these markets closer to guesswork than underwriting.

What this means for a buyer evaluating a site right now

Treat right-to-compute statutes as a possible future tool, not a present-day shield. If your target state has passed or is actively considering one — currently Montana, with New Hampshire, Ohio, and South Dakota in the pipeline — it’s worth tracking closely, but it should not change your underwriting timeline until a court actually rules on a case like Missoula’s. A statute with no precedent behind it is a legal argument, not a guarantee. Separately, map your target jurisdiction’s restriction by type, because “moratorium” covers wildly different levels of risk: a 45-day pause while a county drafts an ordinance (as in Imperial County, California, on its second such pause) is a very different underwriting problem than a permanent ban already on the books in Walton County, Florida, or an 18-month freeze running into 2028 in Cary, North Carolina. Build schedule contingency around close votes, not just around utility interconnection dates — Memphis is a reminder that a single council member’s attendance can be the entire difference between a project proceeding on schedule and losing a year and a half. And keep watching the state legislative layer as closely as the local one: with 24 states carrying pending bills and federal legislation like the Ratepayer Protection Act still alive in some form, the rules governing a site you option today can look materially different by the time you’re ready to break ground, in either direction. This is the same discipline we’ve argued for in our earlier look at community opposition as a site-selection risk, except the fight has now escalated from town halls and council votes into actual courtrooms, with a brand-new legal theory that nobody can yet say will survive contact with a judge.

The practical takeaway for anyone deploying capital into new capacity this quarter is simple: don’t bet a project’s timeline on a right-to-compute law bailing you out of a local restriction, and don’t assume a moratorium that failed once will stay failed. Underwrite the specific restriction in front of you, its type and expiration date, and the political margin behind the vote that created or defeated it — and build the legal uncertainty of untested statutes into your contingency, not your base case. For a structured way to model regulatory, legal, and site-selection risk before committing capital, see the TechInfraHub Data Center Buyer’s Toolkit.

Written by

Raajeev Ratra

Data Center Infrastructure Expert | 21+ Years in DC Design, Operations & Project Management

Raajeev is a seasoned data center professional with hands-on experience in hyperscale facilities, colocation design, power & cooling infrastructure, and global DC operations. He shares practical insights to help engineers and IT leaders build better infrastructure.

Connect on LinkedIn →