‘Pay dividend to community’: Donald Trump tells companies as backlash against data centers grows across US

A data center is a facility filled with computer servers, storage systems, and networking equipment. It processes, stores, and transfers digital information. Artificial intelligence needs powerful computing infrastructure to train models and deliver AI services. That is why companies are expanding data-center capacity.

The article describes these facilities as the infrastructure fueling the rise of AI. More than 3,000 data centers already operate in the United States, while over 1,500 new facilities are in development. Their size can make them major users of electricity and other community resources.

This growth has created a political flashpoint. Communities worry that new facilities may receive priority access to electricity and water. Planned construction is concentrated in the South and Midwest, which together will receive three-quarters of planned centers. Opposition has already blocked or delayed $64 billion in projects.

A community dividend means a company gives money to the community, or to individuals living there, when it operates a data center. The payment is intended to share some financial benefit with people who live near the project. Trump proposed the idea because local residents often bear concerns about large facilities without feeling they receive enough in return.

Trump said data centers that were once highly unpopular in an area became extremely popular after a financial deal was reached. Under his suggestion, companies could pay a community or individual residents a dividend. The article does not specify the payment amount, schedule, or legal structure.

Such payments could change the local debate by making the project financially useful to residents. However, the article also describes concerns about water, electricity, and energy costs. A dividend could improve popularity, but it would not necessarily resolve every infrastructure or planning issue.

The planned expansion is large. The article reports more than 3,000 operational data centers across the United States. It also says over 1,500 additional facilities are in various stages of development nationwide. The figure shows that the current buildout extends well beyond a handful of individual projects.

Construction will be concentrated geographically. Three-quarters of all planned data centers are expected in the South and Midwest. The South alone accounts for nearly half of planned facilities, at 48 percent. These projects are tied to the computing infrastructure supporting artificial intelligence.

The expansion is meeting organized resistance. In July, 142 protests took place across 42 states on one day. Data Center Watch reported that $64 billion in US data-center projects had been blocked or delayed. Public opposition could therefore affect how quickly the planned expansion proceeds.

Data centers can be very large industrial facilities, so communities worry about how much electricity and water they will require. Residents fear that companies building AI infrastructure could receive priority access to these resources. That concern matters especially where local supplies, grid capacity, or utility systems are already limited.

The article links rising data-center demand directly to household cost fears. In a June Reuters/Ipsos poll, 77 percent of respondents said they worried that surging energy needs would drive up electricity prices. The same poll found 57 percent opposed new data centers in their communities. These figures show that opposition involves both local impacts and broader bills.

The concern is influencing projects. In July, protests occurred across 42 states, and Data Center Watch said $64 billion in projects had been blocked or delayed. The article does not quantify water use, but it identifies water and electricity as key resources driving local resistance.

When a data center generates its own electricity, it produces power for its operations rather than depending entirely on the local electricity grid. Selling or giving excess power back means unused generation can support nearby users or be transferred to the wider grid. This could reduce some pressure on shared supplies.

Trump said large data centers have been mandated to make their own electricity. He also said they must give or sell any surplus to the local community or grid. The article does not identify the technology used, the size threshold, or how the requirement is enforced.

The proposal connects data-center growth with community concerns about electricity. Residents fear that AI facilities could increase energy demand and household costs. A self-generation requirement may address part of that concern, but the article does not say whether it has prevented opposition or reduced project delays. Resistance remains widespread, with $64 billion in projects reportedly blocked or delayed.

Opposition has crossed ideological boundaries because the concerns are practical and local. Residents worry about access to water and electricity, possible increases in power bills, and the rapid pace of construction. Those issues can affect people regardless of whether they usually support the same political parties or policies.

The article describes data-center opposition as uniting voters across ideological lines. A Reuters/Ipsos poll found 64 percent disapproved of the fast construction pace, while 57 percent opposed new centers in their own communities. In July, 142 protests took place across 42 states during one nationwide effort.

The resistance is producing concrete results. Data Center Watch reported that $64 billion in US data-center projects had been blocked or delayed by growing local, bipartisan opposition. As more than 1,500 projects remain in development, local acceptance, resource concerns, and planning decisions may continue shaping where construction proceeds.

Electricity grids and water systems set practical limits on new industrial development. A grid must have enough generation and transmission capacity, while water systems must have adequate supply and treatment capacity. If a project could strain either system, residents and officials may question whether existing users will face higher costs or reduced access.

Local planning decisions turn those limits into rules. Officials can assess a proposal, consider its resource demands, and decide whether it fits local plans and regulations. The article shows why this matters: residents fear AI infrastructure will receive priority for electricity and water, while opposition has blocked or delayed projects.

The current expansion is testing these systems across the country. More than 1,500 facilities are under development, with three-quarters planned for the South and Midwest. The article does not provide technical capacity figures for specific grids or water systems. It does show that local approval and resource concerns will shape the buildout.