The Digital Press

AI’s rapidly growing use is driving a surge in energy consumption from data centers, with emissions and infrastructure demands largely underreported and underestimated. While individual AI queries use modest energy, the expanding scale and complexity of AI applications could soon reshape global energy grids and carbon footprints.

Why it matters: AI’s energy needs are skyrocketing, potentially consuming as much electricity as 22% of U.S. households by 2028, with increasing reliance on carbon-intensive sources.

The big picture: Data centers have doubled energy use since 2017 due to AI hardware demands, reversing prior efficiency gains and prompting massive infrastructure investments.

The stakes: Lack of transparency from AI companies hinders accurate emissions tracking and risks shifting energy costs to consumers through utility rate increases.

Commenters say: Many stress the need for open measurement and accountability over AI’s energy use, noting that individual query impacts are small but aggregate demand and corporate secrecy pose serious challenges.