The municipality of Austin, one of America's key technology hubs, has initiated the process of developing regulatory restrictions for new data centers. The reason is growing pressure on the region's resource base, particularly on freshwater reserves and power grid capacity. This decision reflects a broader trend: local authorities across the country are beginning to reconsider the terms for hosting energy-intensive infrastructure.

Austin joins a number of cities in Texas, as well as localities in North Carolina, Kentucky, and New Jersey, where bans or temporary moratoriums on the construction of such facilities have already been introduced. The situation is exacerbated not only by rising demand for computing power from the AI sector, but also by objective environmental consequences.

My calculations and analysis of industry data show that the scale of the problem is significant. Power plants supporting data centers in seven U.S. states consume approximately 3.4 trillion gallons of freshwater annually—comparable to the yearly water consumption of major metropolises. Moreover, at the start of this year, at least 75 projects with a total value of around $130 billion faced organized opposition from local communities and activists.

It is telling that this is not about isolated protests, but a systemic shift in the perception of digital infrastructure. Previously, data centers were seen as an unequivocal boon for the economy; now they are viewed through the lens of resource trade-offs. In Austin, where climatic conditions already strain the water system, this issue is particularly acute.

My expert assessment: This trend is merely the tip of the iceberg. In the coming years, we will see not just restrictions, but a transition to requirements for mandatory compensation of resource costs. Data center operators will have to invest in closed-loop water supply systems and renewable energy, otherwise their projects will become economically unviable in a number of regions. This will reshape the map of global computing infrastructure deployment.