Gov. Hochul Responds to Criticism, Explains Rationale for AI Data Center Moratorium
New York Gov. Kathy Hochul had an op-ed published in The Wall Street Journal on July 21 responding to the publication’s criticisms of her decision to halt data centers of a certain size from being built in her state.
In her op-ed, Hochul explained her stance on AI data centers and why she believes her decision to prevent the creation of large AI data centers in the state for a year was best for New Yorkers. The moratorium on the facilities is the first of its kind in the nation.
She was countering a July 14 WSJ editorial titled “New York’s Data Center Self-Sabotage” published on the same day that Hochul signed her executive order for a statewide moratorium on large data centers, with the newspaper’s editorial board calling her move “an act of monumental self-sabotage.”
The editorial board said the moratorium on the centers was a popularity move more than a pragmatic one, and that the centers helped local areas with tax money. They also said that the issues of data centers taking up utilities like electricity are overstated.
The publication expressed suspicion of the one-year moratorium, comparing it to the 2008 one-year moratorium on fracking in the state, which was extended until fracking was entirely banned years later.
Hochul opened her July 21 op-ed response by saying the WSJ editorial had gotten New York’s decision and rationale for the AI data center moratorium wrong.
She said the one-year pause on hyperscale AI data centers that use 50 megawatts or more is a responsible move, allowing for the establishment of rules for an industry that will hold much influence in the future.
According to Synergy Research Group, these large-scale data centers are the minority of operational centers, but are becoming more common, with hyperscale data centers doubling between 2019 and 2024.
AI data centers are different from regular data centers, which handle running software, web pages, and storing files. AI centers require different computer components focused on AI operations and require more energy and cooling.
