Meta Platforms is accelerating one of the most aggressive data center buildouts in corporate history, pouring tens of billions of dollars into sprawling campuses designed to power its next generation of artificial intelligence systems. The company has announced or begun construction on facilities across multiple U.S. states, with individual campuses sometimes spanning hundreds of acres and drawing enough electricity to rival small cities. This infrastructure arms race is driven by the computational demands of training and running large language models, AI-generated content systems, and recommendation algorithms that underpin Facebook, Instagram, and WhatsApp.
The scale of investment is staggering even by Silicon Valley standards. Meta has committed to spending more than $60 billion on capital expenditures in a single fiscal year, with the bulk directed toward AI infrastructure. Executives including CEO Mark Zuckerberg have framed this spending as existential — a necessary bet that the company that controls the most powerful AI infrastructure will define the next decade of the internet. Data centers are now the physical backbone of that ambition, and Meta is racing against rivals including Google, Microsoft, and Amazon to secure land, power contracts, and construction capacity.
To attract these facilities, states and municipalities have long offered data center operators generous tax incentives, particularly exemptions on sales taxes for equipment purchases. Because a large data center can require hundreds of millions of dollars in servers, cooling systems, and networking hardware, those exemptions can represent enormous sums of foregone public revenue. Virginia, Texas, Iowa, Georgia, and several other states have enacted or expanded such exemptions specifically to compete for hyperscale data center investment, arguing that the facilities create construction jobs, property tax revenue, and long-term economic anchors.
But a backlash is building. Local governments near Meta's facilities in states including Iowa and Texas have begun scrutinizing whether the promised economic benefits materialize at the scale originally projected. Critics point out that fully operational data centers employ surprisingly few permanent workers — often only dozens of full-time staff to manage facilities covering millions of square feet — while placing significant demands on local power grids, water supplies used for cooling, and road infrastructure. Community groups and some local officials argue that the math of tax exemptions does not pencil out when weighed against these costs and against the displacement of other potential land uses.
Several state legislatures are revisiting or tightening their incentive frameworks. Some are adding employment thresholds that companies must meet to retain exemptions, while others are introducing clawback provisions tied to power consumption and environmental impact. The renegotiation of these deals is creating friction with Meta and other hyperscalers that structured their investment decisions around the original incentive packages.
Beyond taxes, data centers have become flashpoints over energy and water. A single large Meta facility can consume hundreds of megawatts of electricity, enough to power hundreds of thousands of homes. In regions where the grid is already strained — including parts of the Midwest and Southeast — utility companies are warning that the rapid addition of hyperscale loads is complicating long-term capacity planning and, in some cases, forcing the extension of fossil fuel generation that might otherwise have been retired. Environmental advocates have pushed back sharply, arguing that tech companies' public commitments to clean energy do not square with the real-time grid realities of their expansion.
Water consumption for cooling is an equally charged issue. Evaporative cooling systems, which are among the most efficient methods for managing the heat generated by densely packed servers, can consume millions of gallons of water daily. In drought-prone areas of the American West and Southwest, where Meta has also explored expansion, this has triggered opposition from agricultural users, municipal water authorities, and conservation groups. The company has invested in more water-efficient cooling technologies, but critics argue the pace of efficiency improvements lags far behind the pace of capacity growth.
At the community level, opposition has taken on a distinctly local character. Residents near proposed or existing facilities have organized against the rezoning of farmland, the visual and acoustic footprint of large industrial campuses, and concerns about property values. In some cases, local elected officials who initially championed data center deals have faced constituent pressure to renegotiate or block further expansion.
The data center tax debate is becoming a larger political issue as AI investment transforms from a niche technology story into a mainstream economic and infrastructure question. Governors and economic development officials who once competed fiercely to land these facilities are now navigating more complex political terrain, balancing the prestige and headline investment figures against voter frustration over strained services and limited job creation. The dynamic is playing out in states that cross partisan lines, suggesting that resistance to unconditioned data center subsidies is not primarily ideological but practical.
At the federal level, the concentration of AI infrastructure investment is drawing attention from both antitrust-minded legislators and national security hawks concerned about the geographic clustering of critical digital infrastructure. Proposals have circulated to create federal standards for data center siting, permitting, and incentives — though no comprehensive legislation has advanced. The Biden-to-Trump policy transition also introduced uncertainty about federal clean energy mandates that intersect with data center power sourcing decisions.
For Meta, the stakes of getting the policy environment right are high. Any significant erosion of state tax exemptions across its primary buildout markets could meaningfully increase the cost basis of its AI infrastructure, affecting the economics of competing with well-capitalized rivals. The company has deployed lobbyists and economic impact studies to defend existing incentive frameworks, but the political momentum in several key states appears to be shifting toward greater scrutiny and conditionality.
Gist is a free AI reader for your browser, iPhone, and Android. Get concise summaries and key takeaways from any article or podcast.
Get Gist — Free