Trump Administration Uses Google AI to Draft Federal Regulations, Raising Safety Concerns

WASHINGTON — The Trump administration is planning to use Google’s Gemini artificial intelligence to draft federal regulations, with the Department of Transportation (DOT) serving as the lead agency for the initiative.

Gregory Zerzan, the DOT’s general counsel, said President Trump is enthusiastic about the plan and wants the department to be the first to "fully use AI to draft legislation."

The primary objective of the move is speed. Zerzan stated that the government does not require a "perfect rule," but rather one that is "good enough" to "flood the market."

Under the new plan, Gemini can generate a draft regulation in less than 20 minutes. The process typically takes months for federal employees to complete.

The move has drawn sharp criticism from experts and former officials who warn that the technology is prone to errors. They cautioned that using AI for critical safety standards—such as aviation or the transport of hazardous materials—could lead to casualties.

One former DOT AI director compared the initiative to "letting a high school intern draft rules."

The Trump administration’s pivot toward using artificial intelligence to draft federal regulations marks a tectonic shift in American governance, signaling a departure from the deliberate, expert-led deliberation of the traditional administrative state. By prioritizing speed and volume over meticulous scrutiny, the administration is effectively importing Silicon Valley’s “move fast and break things” ethos into the federal bureaucracy—a domain where errors translate into public safety risks rather than mere software glitches.

The strategy of “flooding the zone,” as articulated by a senior Department of Transportation official, suggests a calculated political maneuver. By saturating the regulatory environment with a massive volume of new mandates at high velocity, the administration seeks to fundamentally reshape the legal landscape before institutional or legal opposition can mount an effective response.

However, the reliance on commercial large language models (LLMs) like Google’s Gemini introduces profound concerns regarding transparency and accountability. The transition of federal lawmaking from human experts to algorithms means that the latent biases embedded in proprietary training data may now be codified into national policy. Furthermore, this shift creates an unprecedented legal vacuum: should an AI-drafted regulation lead to a systemic failure or a public safety incident, the challenge of apportioning liability between the federal government and technology providers remains a complex and untested frontier in administrative law.

The impact on the Vietnamese-American community aligns with that of the general public, focusing primarily on transportation and traffic safety regulations. These measures affect the daily lives of everyone in the community, whether you're commuting through Little Saigon or driving to a local nail salon or phở restaurant. Ultimately, these regulations ensure that road safety remains a practical, everyday priority for all residents.