Ghost Data Centers Surge in US as AI Power Demands are Inflated

Introduction

In recent months, the United States has witnessed a troubling trend: companies submitting plans for data centers that may never actually be built, all in an effort to secure vast amounts of electricity for future artificial-intelligence (AI) workloads. These “ghost data centers,” often lacking credibility or viable backing, are rapidly distorting the perceived demand for power. As a result, regulators and energy providers are scrambling to adjust infrastructure planning and pricing models — and the ripple effects may strain both markets and communities.

Original Report

The phenomenon centers around firms proposing numerous data center construction projects – often in the American Midwest – ostensibly to support compute-intensive AI services. These proposals are presented to utility companies and power providers not to immediately draw power, but to reserve it well in advance. By exaggerating projected energy demands, these developers aim to secure electricity capacity early on. Observers say many of these proposals lack concrete timelines or funding, raising doubts about their feasibility.

The strategy grants companies leverage: once they lock in power, they retain the option to build later — or never. As a result, the aggregate reported demand for electricity inflates significantly beyond actual use. Estimates suggest that more than 60 percent of these “paper plans” may disappear during payment-verification or follow-up phases. Meanwhile, the bloated demand distorts utility planning for new power plants and disrupts fair pricing models.

Government agencies have started voicing alarm. They argue the speculative filings risk undermining both the stability of supply chains and the fairness of rate-setting mechanisms. The surge in ghost filings coincides with the rapid growth of generative‑AI platforms, like conversational AI and image generation tools, which rely on heavy computing power. As global attention focuses on regulation and responsible growth of AI, critics argue such practices reflect a speculative bubble, not genuine infrastructure development.

What Undercode Say:

The surge in ghost data center proposals is symptomatic of deeper structural flaws in how energy demand and infrastructure planning are tied to speculative corporate intent. At face value, securing power in advance may seem like prudent risk management. In reality, it distorts resource allocation — inflating perceived demand, triggering overinvestment in generation capacity, and potentially saddling ratepayers with higher costs.

First, utility providers base generation investments, grid upgrades, and pricing tiers on forward-looking data. If a substantial portion of that data is based on speculative or non‑viable projects, design decisions may overcompensate. This can lead to a mismatch where expensive power plants or grid expansions are justified — even though the actual load never materializes. The sunk costs in infrastructure might then be recuperated through increased rates or fees, indirectly burdening consumers.

Second, from a regulatory standpoint, the ease with which companies can submit unverifiable proposals points to lax submission and verification standards. The fact that 60 percent or more drop out during later validation suggests that initial scrutiny is too weak. This opens the door not only to speculative gaming, but also to potential misuse — such as hoarding of capacity, market manipulation, or attempts to lock favorable rates ahead of demand surges.

Third, on the AI‑industry side, linking data center construction to generative‑AI hype creates a dangerous echo chamber. Investors and developers may exaggerate demand projections to attract funding or favorable utility contracts — even if their actual AI workloads never reach the projected scale. In effect, “AI growth” becomes a mythic justification rather than a business case. The risk is a massive overhang of stranded or underused infrastructure.

Furthermore, the consequences are not just economic. Overbuilding power supply in anticipation of phantom demand could lead to unnecessary environmental impacts if new plants are constructed, increased energy waste, and inefficient resource distribution. It shifts the burden from corporate risk management to public good — misusing collective infrastructure to hedge individual speculative bets.

Finally, there is a reputational risk to the broader AI ecosystem. As regulators and the public grow wary of inflated claims, legitimate AI operators may face more skepticism, stricter scrutiny, and delayed approvals — slowing down real, valuable innovation.

Fact Checker Results:

✅ Reports show a substantial number of data‑center power reservations in the U.S. appear speculative or unverifiable.
✅ Utility and regulatory bodies have noted the distortive impact of over‑inflated power demand on infrastructure planning and pricing.
❌ There is no public evidence that all—or even most—of the proposed data centers will be built as originally planned.

Prediction

⚡ In the next 12 to 24 months, expect a substantial contraction in announced data‑center projects in the U.S. as regulators tighten verification requirements and utilities demand firmer proof of use. Many ghost plans will be canceled or delayed. Grid upgrade proposals based on inflated demand may be scaled back or postponed. AI infrastructure investment could decelerate, pushing firms to favor efficiency and reuse over speculative expansion.

🕵️‍📝✔️Let’s dive deep and fact‑check.

References:

Reported By: xtechnikkeicom_f0bc4d61731b9128e1d96ab0
Extra Source Hub (Possible Sources for article):
https://www.github.com
Wikipedia
OpenAi & Undercode AI