US Department of Transportation Embraces AI for Federal Rule Drafting
The Trump administration is reportedly advancing a significant initiative to incorporate artificial intelligence into the federal rulemaking process. According to recent reports, the US Department of Transportation may utilize AI technologies to draft some federal transportation regulations. This move represents a broader federal effort to accelerate regulatory processes through technological innovation.
AI Demonstration and Leadership Enthusiasm
In January, DOT staff attended a demonstration showcasing AI's potential to revolutionize rule drafting. Agency attorney Daniel Cohen communicated to colleagues that the event would highlight exciting new AI tools available to DOT rule writers to enhance efficiency and speed. Meeting notes reviewed by ProPublica indicate that discussions about this plan continued among agency leadership recently.
During a leadership meeting, Gregory Zerzan, the agency's general counsel, stated that President Donald Trump is very excited about this initiative. Zerzan positioned DOT as leading a broader federal effort, describing the department as the point of the spear and the first agency fully enabled to use AI for drafting rules. His comments suggested a focus on volume over perfection in regulation production, emphasizing the goal of flooding the zone with regulations.
Concerns About Safety and Accuracy
Some DOT members have expressed concerns about delegating the writing of vital transportation safety standards to emerging AI technology. The agency's regulations cover critical aspects like airplane safety, gas pipeline explosions, and freight train derailments. Staffers questioned why such important standards would be entrusted to technology known for generating errors.
However, supporters of the plan emphasize speed as the primary advantage. Drafting and revising complex federal regulations typically require months or even years. With tools like Google Gemini, employees could potentially draft proposed rules in minutes or seconds, according to staffers who attended demonstrations. Presenters suggested that much of the content in regulatory document preambles is merely word salad that AI can handle effectively.
Accelerated Timeline and Implementation
Zerzan reiterated the goal of speeding up rulemaking with AI during recent meetings. The objective is to shorten the timeline for producing transportation regulations, allowing concepts to become complete drafts ready for review within just 30 days. He suggested this should be achievable because it shouldn't take more than 20 minutes to get a draft rule from Gemini.
These plans are already in motion. The department has employed AI to draft a still-unpublished Federal Aviation Administration rule, according to a DOT staffer briefed on the matter. This initiative opens a new chapter in the administration's effort to incorporate AI into government operations, building on previous executive orders and action plans supporting AI adoption.
Criticism and Support from Experts
Critics argue that large language models like Gemini and ChatGPT shouldn't be entrusted with complex governance responsibilities due to their error-prone nature and lack of human reasoning. However, advocates view AI as a means to automate routine tasks and improve efficiency within the federal bureaucracy.
This optimistic perspective was evident at recent AI summits where federal technology officials discussed adopting an AI culture in government. Justin Ubert, division chief for cybersecurity at DOT's Federal Transit Administration, participated in panels about fast AI adoption. He noted that many view humans as a choke point slowing down AI and predicted that eventually humans would step back into oversight roles monitoring AI-to-AI interactions.
Staff Reactions and Practical Demonstrations
During a December presentation that drew over 100 DOT employees, presenters demonstrated how Gemini could handle 80% to 90% of regulation drafting work. The presenter asked for a topic requiring a Notice of Proposed Rulemaking, entered keywords into Gemini, and generated a document resembling such a notice. However, staffers noted it appeared to miss actual regulatory text.
Attendees said the presenter showed little concern about AI hallucinations erroneous text frequently generated by large language models. He suggested DOT staff would proofread machine-generated products. Some employees recalled the presenter as Brian Brotsos, the agency's acting chief AI officer.
Expert Warnings and Workforce Concerns
The presentation left some DOT staffers with serious doubts. They emphasized that rulemaking requires expertise in subject matter, statutes, regulations, and case law. Errors in DOT regulations could lead to lawsuits, injuries, or deaths in transportation systems. Some rule writers have decades of experience that seemed overlooked by AI proponents.
Mike Horton, DOT's former acting chief AI officer, criticized the plan, comparing it to having a high school intern do rulemaking. He revealed the plan wasn't in development when he left in August 2025 and warned that going fast and breaking things means people could get hurt.
Academics monitoring AI in government expressed varied opinions. Bridget Dooling, an administrative law professor at Ohio State University, suggested AI could be beneficial as a research assistant with substantial supervision but warned against delegating too much responsibility. Ben Winters of the Consumer Federation of America expressed concern given the departure of subject-matter experts from government due to workforce cuts.
Broader Context and Administration Position
Elon Musk's Department of Government Efficiency has been a strong advocate for AI adoption in government. A leaked DOGE presentation from July 2025 proposed using AI to eliminate half of all federal regulations by automating document drafting. The presentation stated that writing is automated and AI automatically drafts all submission documents for attorneys to edit.
The White House hasn't confirmed whether similar AI rulemaking plans exist at other agencies. Four senior technology officials said they weren't aware of such plans. Regarding DOT's claim of being the point of the spear, some officials expressed skepticism, suggesting it might be more about marketing than actual leadership in federal AI adoption.
