Why Should Texans Pay for Billion-Dollar Data Centers Without a Clear Return? The AI Boom, Tax Breaks and the Grid Bill Nobody Asked For

Imagine a corporation announces an enormous data center in Texas. There is talk of artificial intelligence, innovation, jobs and billions of dollars in investment. The ribbon-cutting photographs look wonderful. But a homeowner down the road has a question that rarely gets answered in those glossy announcements: If the electrical grid has to be expanded for this project, how much of the bill ends up on my utility statement?

That is not hostility toward technology. It is a basic question about who pays and who benefits. Texas has positioned itself as a major destination for data centers, and the industry is expanding fast. Yet the public debate increasingly involves water, land, transmission lines, tax incentives and whether residential customers should shoulder costs created by some of the largest electricity users in the state.

Let's establish what a data center actually does. It houses computer equipment used to store information, run applications, provide cloud services and support artificial intelligence. These facilities are vital to many digital services. But unlike a small office, a large data center may require enormous quantities of electricity, specialized cooling and major infrastructure. It can operate around the clock even when the surrounding neighborhood is asleep.

The scale of proposed growth has become difficult for ordinary consumers to comprehend. In August 2026, the Texas Tribune reported nearly 2,000 requests to connect large users to the ERCOT grid, approximately 90 percent associated with data centers. The combined requested demand was reported at 466,497 megawatts, more than five times the grid's historical peak demand. That is a queue of proposals, not a measurement of electricity already consumed. Many projects may be speculative, delayed or never completed. But even a fraction of that demand would require serious planning.

Why do these proposals matter to someone who has never set foot in a data center? Because electricity does not travel through wishful thinking. A project may need substations, transmission lines, generation resources, grid studies and upgrades to existing infrastructure. Somebody must pay to build and maintain those assets. Regulators decide which costs are directly assigned to the project and which may be shared across customers.

Shared infrastructure is not automatically a scam. A new transmission line can serve multiple communities, support industrial growth and improve reliability. It may also reduce congestion costs over time. But the benefits are not always distributed evenly. When a particular private facility is the main reason an expensive upgrade is required, residents have every right to ask whether the private customer is paying its fair share.

Here's where the conversation gets really interesting. In June 2026, Governor Greg Abbott directed the Public Utility Commission of Texas and ERCOT to protect residential customers from data-center infrastructure costs. His office specifically called for data centers to fully fund the electric infrastructure required for their operations. He also urged action to lower residential transmission costs and proposed future legislation on water efficiency, disclosures and outdated tax incentives. The governor's intervention is strong evidence that the cost-allocation issue is real enough to command statewide attention.

But a directive is not the same as a completed audit of every bill. Consumers should ask what rules have been adopted, how they apply to existing and proposed facilities, what happens when infrastructure benefits multiple users, and whether any costs are still recovered from residential customers. A promise to protect ratepayers is the beginning of accountability, not its conclusion.

Now let's discuss the tax breaks, because this is where the public return-on-investment question becomes especially sharp. At an August legislative hearing, Texas Tribune reporting said the state comptroller estimated that the data-center sales-tax exemption would amount to approximately $1.3 billion in forgone state sales taxes in 2026. That is a tax expenditure estimate, not a direct check written to one company, and it does not mean every dollar would necessarily have been collected if the exemption never existed. But it is a substantial public policy choice.

Supporters of tax incentives argue that they attract capital investment, construction activity, local property tax revenue, jobs and business growth that might otherwise go to another state. Critics question whether the benefits justify the lost revenue, especially when facilities can be highly automated after construction and may employ fewer permanent workers than their enormous buildings suggest. The honest answer requires project-level evidence: how many permanent jobs, at what wages, for how long, and what taxes are actually paid after exemptions?

Imagine a family business asking the state to waive a large portion of its taxes because it promises to create jobs. Policymakers would presumably want numbers, timelines and accountability. The same standard should apply when the applicant is a technology giant or a developer building infrastructure for one. Size should not substitute for proof of public benefit.

Then there is water. Many data centers use cooling systems that can consume significant quantities of water, although consumption varies greatly with technology, location and operating design. In water-stressed communities, a project can raise concerns even if it pays its electricity bills. Texans have been asking whether industrial growth will compete with residential water needs and whether communities have meaningful input before projects are approved.

In September, ERCOT issued a request for information about the state and community impacts of data centers, working with the Public Utility Commission and the Texas Water Development Board on questions about water sources and consumption. That process illustrates how much officials still need to understand about the proposals. A company should not be able to claim negligible community impact without disclosing the data needed to test that claim.

The siting issue is especially complicated in rural Texas. Landowners have protested proposed 765-kilovolt transmission corridors that could cross farms, ranches and private property. At an August House hearing, lawmakers heard hours of testimony about the proposed lines, local authority and the tension between infrastructure expansion and property rights. Some residents fear losing control over their land. Energy industry representatives warn that delaying needed lines could undermine reliability and other economic development.

Those competing arguments deserve serious attention. If regulators approve transmission routes too quickly, affected residents may have inadequate opportunities to challenge them. If regulators halt necessary upgrades indefinitely, congestion, reliability risks and future electricity costs may worsen. A credible policy must address both concerns instead of pretending there is a cost-free choice.

Another complication: not every large transmission project exists solely for data centers. Texas also has growing population centers, oil and gas activity, factories, renewable generation and other industrial demand. Assigning every proposed power-line dollar to AI would be inaccurate. But refusing to disclose how much incremental demand comes from data centers would be equally unsatisfactory.

In August, ERCOT officials described an audit covering approximately 250 to 300 large projects, most of them data centers. The projects under review represented about 200 gigawatts of potential demand. The purpose was to verify whether proposals were realistic and what they would require before allowing them to proceed through the connection process. This matters because a speculative queue can distort planning if developers reserve capacity for projects that will never materialize.

Think about the practical consequences. If a developer requests power for an enormous campus that never gets built, planners may still spend time and resources evaluating it. If the request contributes to a decision to expand infrastructure, consumers and landowners may be affected. Better project verification, deposits, cost responsibility and milestones could help distinguish serious development from speculative land grabs for grid access.

So what would a genuine return on investment look like for the public? Start with transparent accounting. For every major project receiving public incentives, disclose the estimated value of those incentives, the permanent jobs created, the wages paid, the local tax revenue, the water demand and the infrastructure costs assigned to the company versus other customers. Publish the actual outcomes later, not just the promises made before construction.

Next, require companies to cover infrastructure costs clearly attributable to their own operations. Where infrastructure is genuinely shared, explain the cost-allocation formula and demonstrate the public benefit. If the public is being asked to subsidize development, identify the benefit it is purchasing. It might be improved reliability, lower costs, tax revenue or economic opportunity. But the answer cannot simply be 'because AI is the future.'

Communities should also have meaningful participation before approvals are finalized. Residents need access to water-use estimates, noise studies, proposed transmission routes and clear descriptions of the tax incentives. Public meetings should occur early enough to influence decisions, not merely after the key agreements have been made.

Some advocates go further and suggest community benefit agreements, workforce-training commitments, local hiring targets or mechanisms that direct part of the economic upside back to affected neighborhoods. Those ideas are proposals, not existing universal requirements. They deserve careful evaluation because benefits should be measurable, enforceable and suited to the community rather than decorative language in a press release.

And here is the uncomfortable truth for the technology industry: if your business model depends on the public absorbing costs while your shareholders capture the profits, people are going to question the arrangement. That does not mean every data center is guilty of cost shifting. It means developers should welcome transparent proof when they say they are paying their own way.

Texas leaders from different political backgrounds have raised overlapping concerns about cost allocation, water use, local control and tax incentives. That cross-party attention is significant. It suggests the debate is not simply about whether somebody likes or dislikes artificial intelligence. It is about how a state manages an enormous industrial transition without ignoring the people who already live there.

As of October 16, 2026, the important distinction is between proposals, directives and verified outcomes. The state has announced protective goals and audits. Regulators are working through large-load connection rules. Lawmakers are discussing tax incentives and infrastructure. None of that establishes that every Texan has already paid a particular data-center surcharge, nor does it prove that the public has received no benefit. The evidence supports a demand for project-specific accounting, not a blanket assumption.

My question for the companies is simple: if this investment is such a gift to Texas, show us the complete balance sheet. Show the jobs, show the taxes, show the water use, show the grid charges and show what households will pay. Then let communities judge whether the deal makes sense.

Because Texans should not be expected to applaud a billion-dollar project merely because it has the letters A and I attached to it. Innovation can be exciting. Accountability should be nonnegotiable.

SOURCES & REFERENCES

Office of Governor Greg Abbott, June 10, 2026, directive on ratepayer costs: https://gov.texas.gov/news/post/governor-abbott-directs-puc-and-ercot-to-shield-texans-from-data-center-infrastructure-costs

Texas Tribune, August 19, 2026, legislative hearing and tax exemption estimate: https://www.texastribune.org/2026/08/19/texas-house-meeting-data-centers-transmission-lines-backlash/

Texas Tribune, August 24, 2026, data centers and transmission concerns: https://www.texastribune.org/2026/08/24/texas-765-transmissions-lines-data-centers-pauses-delays/

Texas Tribune, August 14, 2026, ERCOT audit: https://www.texastribune.org/2026/08/14/texas-data-center-approval-pause-ercot-power-grid/

ERCOT, September 14, 2026, community impacts and water request: https://www.ercot.com/services/comm/mkt_notices/M-B091426-01

ERCOT, June 18, 2026, Batch Zero process: https://www.ercot.com/news/release/06182026-puct-approves-ercots