Opinion | The Data Center Conundrum | Enterprise cmms software
By Gammatek ISPL, Industrial Systems & Compliance Analyst at Gammatek ISPL
Last updated: September 2026 | 14 min read
Author block: Gammatek ISPL advises manufacturing, chemical, and pharmaceutical plants on facility compliance, safety systems, and infrastructure operations at Gammatek ISPL. This piece draws on publicly reported data center policy developments through August 2026 and Gammatek's direct experience with industrial facility compliance and uptime systems.
Why This Matters Right Now
For the past three years, the story of the AI boom has been told almost entirely through chips and models. The physical infrastructure underneath that boom — the data centers actually running the training and inference workloads — has quietly become the bigger constraint, and in 2026 it turned into open political conflict. New York became the first US state to impose a moratorium on new large-scale data centers. Seattle's city council voted unanimously to pause new projects. A California city banned them outright by public vote. If your business depends on cloud infrastructure, enterprise software uptime, or facility operations of any kind, this isn't a distant policy story — it's a preview of cost, capacity, and reliability pressure that's about to work its way through every layer of enterprise IT and facilities management, including yours.
The Conundrum, Defined
The tension is straightforward to state and difficult to resolve: the AI industry needs enormous, fast-growing amounts of data center capacity to keep up with demand, while the communities and power grids hosting that capacity are increasingly unwilling — or unable — to absorb it without real cost. Both sides of that sentence are getting harder, not easier, as 2026 goes on.
On the demand side, the numbers are large enough to strain planning assumptions built just a few years ago. Global data center electricity and water consumption is projected to roughly double by 2030, according to a United Nations report released in June 2026. Worldwide, data centers drew 448 terawatt-hours of electricity in 2025, with AI workloads responsible for about a fifth of that total, and the same facilities consumed 4.5 trillion liters of water and produced 189 million tons of carbon dioxide emissions over the same period. In the US specifically, large data center power demand climbed by roughly 11.3 gigawatts in 2025 alone, reaching 61.8 GW — and forecasts from 451 Research put total demand for IT load, cooling, and lighting at 75.8 GW in 2025, rising to 108 GW by 2028 and 134.4 GW by 2030.
On the resistance side, that growth is now running directly into local politics. Fourteen US states have introduced legislation in 2026 to pause or restrict new data center construction, and more than 100 localities have considered similar measures. New York's moratorium, signed by Governor Kathy Hochul in August, freezes permitting for any new facility requiring 50 megawatts or more while the state conducts an environmental impact review — she's also pursuing legislation to repeal sales tax exemptions that made New York attractive to hyperscale developers in the first place. Seattle's city council moved even faster, voting unanimously for a one-year emergency moratorium after data center proposals under review reportedly carried a combined peak power demand equal to roughly a third of the city's average daily consumption. Monterey Park, California, went further still, becoming the first US city to ban data centers permanently by public vote.
Public opinion backs the resistance more than the industry might prefer: according to a recent Reuters/Ipsos poll, only one in three Americans approve of the current pace of data center construction, and most would oppose one being built in their own community.
Original Analysis: Comparing the Four Major Approaches
Not every jurisdiction is handling this the same way, and the differences matter for anyone trying to forecast where capacity — and cost — will land next.
Jurisdiction | Approach | Threshold | Duration | Stated Rationale |
New York (state) | Executive order freeze | 50 MW+ | Up to 1 year | Grid strain, water use, utility cost impact |
Seattle | Emergency city moratorium | Large-load facilities | 1 year | Power grid capacity, local utility rates |
Monterey Park, CA | Permanent ban (public vote) | All new facilities | Indefinite | Community and environmental opposition |
Maine | Legislative pause (vetoed) | 20 MW+ | Proposed until 2027 | Grid reliability, ratepayer protection |
A pattern emerges once you line these up: the jurisdictions moving fastest aren't primarily driven by ideological opposition to AI — they're driven by utility rate and grid reliability math that local officials can no longer avoid addressing publicly. New York's own language ties the moratorium directly to utility bills, not AI policy in the abstract. That distinction matters for forecasting: this isn't a backlash likely to fade with the news cycle, because the underlying grid capacity math doesn't resolve itself just because political attention moves elsewhere.
What the Industry Gets Wrong About This Debate
A common industry response frames the moratorium wave as anti-technology sentiment or NIMBYism dressed up in environmental language. That framing understates the actual mechanism at work. Data centers are large-load customers whose power draw shows up directly in the rate base calculations utilities use to set prices for everyone else on the grid — meaning nearby residential and commercial customers can see real electricity bill increases tied to data center growth in their region, independent of their own usage. That's a materially different situation than typical industrial development opposition, and it's why the coalition opposing new data centers spans political lines in a way most infrastructure fights don't.
The Layer Almost No One Is Covering: What's Actually Running Inside These Facilities
Most coverage of this conundrum stops at power politics. What gets skipped is the operational reality of running a facility at this scale once it's built — and this is where the story connects directly to enterprise infrastructure decisions far beyond hyperscale campuses.
A modern data center, regardless of size, depends on the same category of operational software any large industrial facility runs on:
Enterprise CMMS software — Computerized Maintenance Management Systems — schedule and track the constant preventive maintenance data center cooling systems, backup generators, and power distribution units require. A single unplanned cooling failure in a high-density AI training cluster can mean six-figure losses in wasted compute time within hours, which is why facility teams increasingly treat CMMS scheduling discipline as a business-continuity function, not a maintenance afterthought.
Enterprise backup and recovery systems protect the actual value sitting inside these buildings — training data, model checkpoints, and customer workloads — at a scale where a single corrupted backup can mean weeks of retraining cost. This is a different tier of requirement than typical corporate backup software use cases, but the underlying category is the same.
Enterprise contract management software handles the dense web of power purchase agreements, colocation leases, water rights arrangements, and increasingly the "bring your own power" deals some operators are now negotiating directly with utilities to bypass grid capacity constraints — a direct operational response to the moratorium wave described above.
Enterprise scheduling software coordinates maintenance windows, staffing, and compliance inspections across facilities that, unlike a typical office building, cannot simply close for a weekend to perform upgrades without risking service-level agreement penalties.
Enterprise property management software increasingly factors into site selection itself, as developers model power availability, water rights, and local regulatory risk alongside traditional real estate criteria before committing capital to a new build.
None of this is incidental detail — it's the actual machinery determining whether a facility survives the tightening regulatory environment described above. A data center with poor maintenance scheduling and weak documentation is the one most likely to draw negative local attention when something visibly fails; one with disciplined operational software and audit-ready records is the one more likely to survive the next round of environmental review.
Implementation Considerations for Facility and Plant Managers
If you manage industrial, manufacturing, or facility operations outside the data center industry directly, this conundrum still has practical implications worth planning around:
Expect regional power cost volatility to spread beyond states with active moratoriums, as utilities reallocate large-load capacity planning nationally in response to political pressure in any single state — plan facility power budgets with wider margins than historical trends would suggest.
Audit-readiness is becoming a competitive advantage, not just a compliance requirement. Facilities that can produce clean maintenance, safety, and environmental impact documentation on demand are better positioned when local review processes intensify — a lesson data center operators are learning under real regulatory pressure right now, and one manufacturing and pharma facilities should apply proactively rather than reactively.
Preventive maintenance scheduling discipline directly reduces regulatory exposure. The same CMMS discipline protecting data center uptime protects any industrial facility from the kind of visible equipment failure that draws local scrutiny and accelerates unfavorable policy attention.
Contract terms with utilities and vendors deserve renewed scrutiny. As "bring your own power" and direct utility negotiation models spread from hyperscale data centers into broader industrial contexts, facilities of any size should revisit their own power and vendor agreements for similar renegotiation opportunities.
Where This Is Headed
The moratorium wave isn't going to resolve the underlying tension — grid capacity, water availability, and local utility economics don't reset just because a one-year pause expires. What's more likely is a slower, more negotiated version of data center growth: facilities that bring their own power generation, operators who invest early in community relationships and transparent environmental reporting, and a widening gap between operators with disciplined facility operations and those without. The political story will keep making headlines. The operational story — who actually has the maintenance, documentation, and compliance discipline to survive tightening scrutiny — is the one that determines who's still building in five years.
That operational discipline isn't unique to data centers. It's the same discipline manufacturing, chemical, and pharma plants need to maintain safety compliance, pass audits, and avoid becoming the visible failure that draws regulatory attention in their own sector.
Comments