Vance Backs AI Data Centers - If Developers Bring Their Own Power

Theresa Payton entered the Google News cycle after JD Vance defended nationwide AI data centers despite a growing fight over electricity bills. The September 1 Fox segment framed the expansion as both an economic opportunity and a national security requirement.

Yet the argument is no longer simply about whether America needs more computing capacity. The dispute now centers on who pays for the power plants, transmission lines, and grid upgrades supporting that capacity.

Vance says developers should bring new generation with them, instead of shifting infrastructure costs to nearby households. That condition separates his defense from President Donald Trump’s blunter demand that communities accept more data centers.

The distinction matters because AI infrastructure has become a local affordability issue. A project can support national computing goals while still creating concentrated costs for one utility territory, watershed, or neighborhood.

Payton’s appearance provided a security-focused response to that conflict. However, the short public description offers less detail than a full policy proposal, leaving important implementation questions unanswered.

What Theresa Payton’s Google News Appearance Actually Added

Payton placed cybersecurity and national resilience inside a debate increasingly dominated by household power costs.

The Fox video listing says Payton analyzed Vance’s defense of nationwide AI data centers on “Fox and Friends First.” The segment appeared on September 1, 2026.

Payton is a former White House chief information officer and the chief executive of cybersecurity firm Fortalice Solutions. That background gives her a different perspective from utilities, developers, or elected officials seeking local tax revenue.

The available description says she addressed the security and economic benefits of America’s continuing infrastructure expansion. It also identifies utility grids and long-term national security as central parts of the discussion.

Those themes are credible areas for analysis, but the available clip summary does not establish a complete policy position. It does not specify which projects deserve approval, which security standards should apply, or how costs should be allocated.

That evidence gap deserves attention. A television response can introduce a useful national security frame without resolving the harder regulatory questions facing states and grid operators.

For AI systems, data centers are physical production facilities. They contain servers, networking equipment, storage systems, cooling equipment, backup power, and connections to the wider electrical grid.

That physical footprint makes AI security broader than protecting software and training data. It also includes electricity availability, equipment supply chains, communications links, facility access, and recovery after an outage.

A country lacking reliable computing infrastructure becomes more dependent on facilities located elsewhere. That dependence can affect commercial cloud services, government systems, scientific research, and military applications.

Domestic capacity therefore carries strategic value. However, capacity alone does not guarantee resilience, especially when facilities cluster around the same congested transmission corridors or depend on scarce equipment.

Security also depends on how operators manage software, contractors, physical access, and hardware procurement. A larger estate creates more capacity, but it can create more attack surfaces as well.

That is the useful contribution from Theresa Payton’s data center analysis. It moves the story beyond a binary choice between unrestricted construction and complete opposition.

The real decision concerns the conditions attached to development. Those conditions determine whether a facility strengthens national infrastructure or transfers new risks to the community hosting it.

The Google News headline compresses that complicated question into a reaction to Vance. The larger story is a test of whether federal ambition can coexist with enforceable local protections.

That test begins with electricity. Security arguments lose public force when residents believe their bills will finance infrastructure built for some of the world’s largest technology companies.

Vance’s AI Data Centers Condition Changes the Argument

Vance defended expansion, but he also acknowledged that higher local electricity bills drive much of the resistance.

Vance spoke after Trump urged communities to embrace data centers and linked the industry to jobs, tax revenue, and competition with China. Trump’s message presented development as a choice between economic participation and falling behind.

When reporters asked about that position, Vance offered a more conditional case. He argued that builders should construct generation alongside new facilities and add electricity rather than remove it from local supply.

According to a published account of Vance’s remarks, he attributed most opposition to areas where projects raise utility costs. His “99 percent” figure was rhetorical and was not accompanied by supporting research.

The condition still matters. It recognizes that national computing capacity and local electricity affordability are connected policy questions.

A data center can require a large, continuous load. The grid must supply that demand during ordinary operations, severe weather, equipment failures, and periods of high household consumption.

New generation does not automatically solve every problem. Electricity must reach the project through substations, transformers, transmission lines, and distribution infrastructure with enough capacity.

The timing also matters. A facility can reach construction milestones before a planned power plant or transmission upgrade enters service, creating an interim burden for the existing system.

Ownership creates another question. If a developer builds generation, regulators must determine who controls it, how it participates in the market, and whether other customers benefit during emergencies.

Vance’s AI data centers position is therefore best understood as a cost-allocation principle. Developers should internalize the infrastructure costs their projects create.

The Trump administration has already tried to formalize that principle through a voluntary Ratepayer Protection Pledge. The expanded ratepayer commitment says large operators should fund the generation and delivery upgrades serving their facilities.

The White House said in July that more than 200 additional utilities, developers, cooperatives, and states had joined. It also said participants represented 80 percent of power delivered to American homes and businesses.

Those are administration figures, not proof that every future project will avoid household costs. The pledge’s effectiveness depends on utility tariffs, regulatory enforcement, project contracts, and transparent accounting.

A voluntary commitment can establish expectations. It cannot replace detailed proceedings before public utility commissions, which decide how many infrastructure costs enter customer rates.

This is where Vance’s condition faces its first practical test. The federal government can accelerate infrastructure, while states retain significant authority over utility regulation and project approval.

The administration’s July 2025 permitting order created a federal definition for covered AI data center projects. It applies to facilities requiring more than 100 megawatts of new dedicated load.

The order seeks faster permitting and greater use of federal land. It also covers supporting components, including transmission equipment, power generation, semiconductors, networking hardware, and storage systems.

That scope reveals the scale of the policy. Washington is not merely encouraging the construction of server buildings.

It is promoting an industrial supply chain that stretches from energy production to advanced chips. Each link carries different permitting, security, financing, and community impacts.

Payton’s national security argument supports that broader view. Vance’s affordability condition determines whether the view can survive contact with local politics.

The Core Tradeoff Is National Capacity Versus Local Exposure

America needs additional computing infrastructure, but national benefits do not erase concentrated local costs.

Supporters see AI data centers as strategic assets. They provide the computing capacity behind model training, cloud services, cybersecurity operations, scientific simulations, and many ordinary digital products.

They also create construction activity and can expand a community’s property-tax base. These benefits help explain why states compete for large projects and offer infrastructure support.

The security argument goes further. Domestic facilities reduce dependence on overseas computing capacity and provide infrastructure subject to American law and operational control.

That advantage matters when governments and businesses process sensitive information. It also matters when international supply chains or geopolitical disputes threaten access to foreign infrastructure.

However, the local balance sheet looks different. Residents experience a project through electric bills, water use, construction traffic, land development, noise, and changes to nearby transmission infrastructure.

Permanent employment can also become a point of contention. Data centers create substantial construction work, but a completed facility does not employ people at the density of many manufacturing plants.

Neither side can resolve this conflict with a slogan. “National security” does not specify a fair electricity tariff, while “local control” does not provide a national computing strategy.

The power numbers explain why the conflict has intensified. A Department of Energy summary of the electricity forecast said data centers consumed about 4.4 percent of U.S. electricity during 2023.

The same research estimated that their share could reach between 6.7 and 12 percent by 2028. That range reflects uncertainty around AI deployment, hardware efficiency, construction schedules, and operating patterns.

Even the lower estimate would represent a major change within five years. Utilities plan generation and transmission over long periods, while AI investment cycles can move much faster.

Grid queues create another complication. New generators, factories, and data centers often seek connections through processes that were not designed for many large requests arriving together.

In June 2026, federal regulators issued orders to the six regional grid operators under their jurisdiction. The connection orders require those organizations to justify or reform rules for data centers and other large loads.

FERC described faster large-load integration as important for innovation and national security. It also emphasized consumer protections, showing that speed and affordability must advance together.

That regulatory action converts the abstract debate into tariff design. Grid operators must decide how projects enter connection queues, demonstrate readiness, fund upgrades, and coordinate with new generation.

Developers want predictable access to power. Utilities want confidence that announced projects will materialize and pay for the infrastructure reserved for them.

Consumers want protection from speculative construction costs. Regulators must balance all three interests without delaying viable projects indefinitely.

This is the primary opponent in the story: the promise of nationally valuable capacity versus the reality of locally concentrated exposure. It is stronger than a simple partisan dispute because it remains regardless of who controls Washington.

Theresa Payton’s data center security case strengthens the capacity side. Vance’s condition implicitly validates the exposure side by conceding that projects become unacceptable when households absorb higher costs.

The two positions can coexist, but only through enforceable project rules. A data center should not receive strategic treatment without accepting strategic obligations.

Those obligations can include dedicated generation, firm capacity, transparent water planning, cybersecurity controls, equipment provenance, and responsibility for grid upgrades.

They can also include performance requirements. If promised generation or community benefits do not arrive, regulators need remedies that do not leave residents carrying the difference.

The Google News treatment presents the moment as a reaction story. The deeper development is the emergence of a conditional political consensus.

Officials can support AI infrastructure while demanding that operators pay their own way. That approach rejects both automatic approval and automatic prohibition.

What the Security Case Does Not Prove

Calling data centers critical infrastructure does not prove that every proposed facility is necessary, secure, or fairly financed.

Payton’s background makes cybersecurity a natural lens for this debate. Still, a security rationale should receive the same scrutiny as claims about jobs, taxes, or consumer savings.

The first uncertainty concerns project purpose. A facility serving sensitive government workloads presents a different national interest from one supporting advertising, entertainment, or speculative capacity.

Most large campuses can host changing workloads over time. That flexibility makes it difficult for local officials to verify claims about a project’s strategic function.

The second uncertainty concerns resilience. Concentrating facilities in regions with cheap power and favorable permitting can produce economic efficiency while increasing geographic dependency.

A storm, fuel shortage, transmission failure, or coordinated cyberattack can affect multiple sites sharing the same infrastructure. More buildings do not guarantee more resilience when they depend on common systems.

The third uncertainty concerns cyber risk inside the facility. Operators must secure management systems, cooling controls, backup generators, physical access, network connections, and vendor support channels.

Operational technology presents particular challenges. These systems control physical equipment, so a compromise can interrupt service even when stored data remains protected.

Hardware supply chains create another layer. Advanced servers contain components sourced through international manufacturing networks that can face shortages, tampering concerns, or export restrictions.

The 2025 federal order recognizes this dependence by treating chips, routers, storage, transformers, and energy equipment as covered components. That definition is a policy acknowledgment, not proof that the supply chain is secure.

The fourth uncertainty concerns electricity accounting. A developer can announce dedicated generation while still relying on the regional grid during maintenance, peak demand, or fuel interruptions.

The project might also purchase contractual energy from a distant source without eliminating congestion near the facility. Financial matching and physical delivery are not always the same.

Regulators therefore need hourly and location-specific information. Annual claims about purchasing enough electricity can hide the periods when a project depends most heavily on shared infrastructure.

The fifth uncertainty is enforceability. The administration’s consumer pledge describes a clear principle, but voluntary commitments can change after a project receives approval.

Long-term tariffs, binding contracts, financial guarantees, and public reporting provide stronger protection. They define what happens if costs exceed forecasts or a promised power project arrives late.

Critics are proposing more direct financial consequences. Senator Ron Wyden’s August 2026 data center proposal would reconsider federal tax incentives and generate support for affected workers and communities.

The proposal represents one side of an active policy debate, not enacted law. Its significance lies in showing that concern extends beyond electricity rates to tax treatment and AI-related labor disruption.

States are testing other approaches, including stricter standards and temporary limits. Those policies can slow investment, but they also reflect the political cost of approving projects before infrastructure rules are settled.

This skeptical angle does not invalidate the security case. It demands evidence connecting each project to a defined public benefit and a credible plan for managing its costs.

That standard should apply equally to supporters and opponents. Developers should document their assumptions, while critics should distinguish between solvable project problems and categorical resistance to new infrastructure.

Vance’s “build the power with the facility” principle is a useful starting point. It remains incomplete until regulators define adequate generation, delivery responsibility, reliability obligations, and consumer safeguards.

Payton’s security framing also needs operational detail. A secure data center requires more than domestic geography or a large capital budget.

It requires tested incident response, segmented networks, reliable backup systems, vetted suppliers, and coordination with public agencies. It also requires enough grid resilience to remain available during a wider emergency.

These requirements increase costs, but they clarify the product being built. The goal is dependable computing infrastructure, not merely the fastest possible addition of electrical load.

Three Signals to Watch After the Google News Cycle

The next phase will be decided by enforceable utility rules, project-level power commitments, and measurable community outcomes.

The first signal is FERC’s large-load reform process. The six regional grid operators must explain whether their existing tariffs handle data centers fairly and efficiently.

Readers should watch for rules governing deposits, project readiness, upgrade costs, connection timing, and coordination with new generation. These details determine whether developers can reserve scarce grid capacity without carrying enough financial risk.

Strong reforms would support Vance’s position by connecting faster approvals to clear payment obligations. Weak or inconsistent rules would leave the same cost disputes to unfold project by project.

The second signal is the structure of new data center power agreements. Announcements should identify the generation source, expected operating date, capacity, ownership, and responsibility for transmission upgrades.

A claim that a project will “bring its own power” means little without those details. The strongest agreements will also explain how the facility operates before dedicated generation becomes available.

Watch whether utilities create separate rate classes for unusually large loads. A well-designed tariff can protect other customers while giving developers predictable terms for long-lived investments.

Also watch for financial guarantees. They can protect customers if a developer cancels a project after the utility begins building expensive infrastructure.

The third signal is evidence from communities already hosting large facilities. Electricity rates, grid reliability, water consumption, tax revenue, construction employment, and permanent jobs provide a more useful record than campaign slogans.

Project-level reporting would let residents compare promises with outcomes. It would also help supportive communities demonstrate when an agreement genuinely works.

These signals will either strengthen or weaken the article’s central judgment. Transparent costs and delivered generation would show that national capacity can expand without shifting unreasonable burdens.

Delayed power projects, rising shared costs, or undisclosed agreements would strengthen the skeptical case. They would suggest that political promises are moving faster than enforceable infrastructure planning.

The Theresa Payton data centers discussion is relevant because cybersecurity depends on these outcomes. A facility with inadequate power planning or disputed community support is not resilient merely because it operates inside the United States.

Public trust is part of infrastructure durability. Projects facing sustained local opposition encounter permitting delays, litigation, political reversals, and pressure for moratoriums.

Developers therefore have a practical reason to exceed minimum disclosure requirements. Clear commitments can reduce uncertainty for residents, utilities, investors, and government customers.

The Vance AI data centers argument will face its sharpest test when a politically important project seeks approval. Officials must decide whether “bring your own power” is a requirement or simply preferred messaging.

Google News will likely continue surfacing the political conflict because data centers now connect AI policy with household affordability. The issue is tangible in a way that model benchmarks and semiconductor specifications are not.

For developers and enterprise buyers, the lesson is straightforward. Computing capacity depends on energy policy, regulatory approval, physical security, and community consent.

For knowledge workers, the connection is less direct but still meaningful. More reliable domestic infrastructure can support the AI services used for research, documentation, analysis, and knowledge management.

However, users should not mistake capacity growth for automatic service improvement. Reliability, privacy, cost, and security depend on how providers design and operate the systems inside those facilities.

The next headline matters less than the contracts and tariffs behind it. Watch whether new projects arrive with operating power, binding safeguards, and public evidence that nearby customers remain protected.

That evidence will determine whether Payton’s security argument and Vance’s affordability condition form a workable policy. Without it, the national data center push remains a promise whose local price is still unsettled.