Thursday, 3 September 2026

Local opposition is mounting against Asia's fast-expanding data center construction, as governments chasing a share of the AI investment boom collide with residents' environmental and health concerns, Semafor Net Zero reported.

The pushback stretches from India to Australia, according to Nikkei. Its effect so far is geographic rather than aggregate: overall demand is not falling, but investment is being redirected as companies move projects toward areas where resistance is weaker, Nikkei reported.

Semafor Net Zero framed the friction as a conflict between the urge of governments to capture value from AI investment and the concerns of their constituents.

Texas has no master mailing list of data center operators, state officials told lawmakers at a state Senate hearing, leaving regulators unable to compile complete figures on how much water and power the facilities consume, according to Inside Climate News.

The gap shows up in the response numbers. The Texas Water Development Board sent its annual mandatory survey to 341 data center operators in January, and by August 26 only about 100 facilities had replied, a response rate of 30 percent, Inside Climate News reported. Reminders and direct messages went to operators who had not answered by July.

What did come back points to a large draw on local supplies. In a joint TWDB survey, 81 facilities reported nearly 1.3 billion gallons of total water intake, of which 964 million gallons was potable water, according to Inside Climate News. Those figures cover a fraction of the operators the state attempted to survey.

The reporting gap sits alongside a hard stop on new grid connections. Governor Greg Abbott essentially imposed a moratorium on new data centers connecting to the Texas grid until an audit can be conducted on the projects' viability and community impacts, Inside Climate News reported.

Officials told the Senate hearing they want to know how much water data centers across the state are using. Without a comprehensive contact list, the survey instrument reaches only the operators the state already knows about.

NextEra Energy has credited an AI-driven dispatch and outage-scheduling tool with saving its customers more than USD 20 million so far this year, according to Power Magazine.

The tool works on two levers utilities watch closely: which units run when, and when plants come offline for maintenance. Power Magazine reported the figure as a customer saving rather than a company earnings item.

Santee Cooper is pursuing a separate use case. The utility expects a custom AI weather-forecasting model to help it avoid spot-market purchases that can cost USD 100,000 an hour during extreme weather, Power Magazine reported.

That framing ties the forecasting work directly to procurement exposure. Sharper weather calls shift the decision of whether a utility covers a demand spike from its own fleet or buys into the spot market at the prices Santee Cooper cited. The company describes the model as custom-built rather than an off-the-shelf product.

Both cases are presented by the utilities themselves in dollar terms, one as realized savings to date, the other as an expected avoided cost.