WASHINGTON — The Trump administration on Thursday advanced measures to curb state-level artificial intelligence regulations, arguing that a patchwork of local rules threatens innovation and imposes ideological requirements on developers. The action targets laws enacted in California, Colorado and other states.

White House officials at the Office of Science and Technology Policy have argued that fragmented state rules create compliance burdens and, in some cases, embed political bias into AI systems through mandated fairness and disclosure standards.

Colorado's AI Act, which imposes obligations on developers of high-risk systems, and California's transparency and safety statutes have been cited by administration officials as examples of overreach that the federal government intends to preempt. The effort revives a debate that surfaced during congressional deliberations over a proposed federal moratorium on state AI laws.

State attorneys general and consumer advocates signalled they would resist any attempt to nullify local protections. Democratic officials in California and Colorado have previously defended their statutes as necessary safeguards against algorithmic discrimination and opaque automated decision-making.

Technology trade groups, including industry bodies representing major AI developers, have lobbied for a single national framework, warning that divergent state rules raise costs. Civil liberties organisations countered that federal preemption without robust replacement rules would leave consumers unprotected against emerging AI harms. Legal experts anticipate court challenges over federal authority to preempt state regulations.