Americans' AI Environmental Concerns Rise Sharply

Americans' concern regarding the environmental impacts of artificial intelligence has significantly increased over the past year, with 53% of adults now reporting they are "extremely" or "very" concerned, according to a new poll conducted by The Associated Press-NORC Center for Public Affairs Research and the Energy Policy Institute at the University of Chicago. This figure represents a notable rise from the 41% who expressed similar levels of concern in the previous year. The poll data indicates that while Democrats are largely driving this surge in environmental awareness, broader anxieties about the local effects of data centers, which are crucial for powering AI and cloud computing, are also becoming more prevalent across the population. Michael Greenstone, director of the University of Chicago institute that collaborated on the survey, commented that "We're undertaking a massive experiment in real time, and I don't think there'll be any putting the milk back in the bottle. Change is scary, and I think some of that is being reflected here." The proliferation of data centers, both smaller conventional facilities and larger hyperscale centers designed to meet the escalating processing demands of AI, is a key factor contributing to these concerns. The United States currently hosts thousands of data centers, with plans for many more to be constructed. Virginia and Texas lead the nation in the number of operational data centers, followed by California, Ohio, and New York. While 87% of existing data centers are situated in urban environments, the trend for new constructions shows a shift, with a greater proportion planned for locations outside of densely populated areas. This geographical distribution of new facilities is a point of interest as the nation grapples with the energy consumption and environmental footprint associated with the rapid advancement and deployment of artificial intelligence technologies. The poll surveyed 1,000 adults in the United States between March 7 and March 11, 2024. The margin of error for the total sample is plus or minus 4.1 percentage points. The findings highlight a growing public awareness of the tangible environmental consequences linked to the digital infrastructure underpinning AI development and its widespread applications.