According to ProPublica reporting, DOT leadership, including General Counsel Gregory Zerzan, has expressed enthusiasm for AI’s ability to rapidly generate rules, emphasizing speed and volume over precision. During a December 2025 demonstration of Google Gemini, agency staff were shown how the AI could produce a draft Notice of Proposed Rulemaking in minutes, though the output often lacked legally required content for the Code of Federal Regulations. DOT leaders reportedly view staff primarily as proofreaders for AI-generated drafts.
Proponents argue that AI could streamline an otherwise slow and complex regulatory process, compressing rulemaking timelines from months or years into days. The initiative aligns with the Trump administration’s broader push to integrate AI across federal agencies, following executive orders, an April 2025 Office of Management and Budget memo, and a June 2025 AI Action Plan. DOT has reportedly already used AI to draft an unpublished Federal Aviation Administration rule. AI advocates within the department envision a future where human oversight is minimal, monitoring “AI-to-AI interactions” rather than performing substantive rulewriting.
However, the plan has sparked concern among DOT staff and experts. Transportation regulations cover safety-critical areas, from pipeline and airplane operation to hazardous material transport, and errors could result in injuries or fatalities.
Critics warn that large language models like Gemini are prone to mistakes, hallucinations, and lack human reasoning, making them risky for high-stakes regulatory work. Former DOT AI chief Mike Horton likened the approach to having a “high school intern” write regulations, while academics such as Bridget Dooling stress that AI-generated text does not guarantee high-quality government decisions. Concerns are heightened by the administration’s workforce reductions, which have left DOT with 4,000 fewer employees, including over 100 attorneys, limiting the agency’s capacity to review AI outputs effectively.
Despite these concerns, DOT leadership remains focused on rapid deployment, positioning the department as the “point of the spear” in federal AI adoption. Critics and some federal officials, however, see this as largely symbolic, a marketing move rather than a carefully calibrated regulatory innovation. As the debate continues, the Trump administration’s AI-driven rulemaking experiment raises urgent questions about balancing efficiency with accountability and safety in critical public infrastructure.
Read the full ProPublica report here.