Democratic US senator Ed Markey of Massachusetts has unveiled a package of bills he calls an "AI accountability agenda", aimed at curbing what he sees as the wide-ranging harms of unregulated artificial intelligence. The agenda targets energy- and water-intensive datacenters, intrusive workplace surveillance, biased algorithms, AI overriding human judgment and deepening economic inequality, framed as an effort to "take power back from big tech". It matters because there has been little federal action on AI since ChatGPT's release in 2022, and Markey argues a piecemeal, state-by-state approach would leave too many people exposed.

Central to the package is a forthcoming bill requiring datacenter owners to obtain certification from the Federal Communications Commission, confirming a facility "will not harm the public interest" before construction begins, with the FCC assessing effects on air and water quality, noise, energy costs, grid reliability, wildlife, jobs and the local economy in consultation with bodies such as the EPA. Other proposals would bar employers from relying primarily on automated systems for hiring and firing, mandate independent bias audits of AI, require human override options in healthcare, and protect workers who disagree with AI recommendations. Markey, 79 and seeking a third full term, is optimistic about eventual bipartisan support; one related measure, the Children and Teens' Online Privacy Protection Act, passed the Senate in March.