Trump Administration Plans to Fast-Track DOT Regulations Using Google Gemini

The intersection of governance and generative artificial intelligence has reached a pivotal new milestone. The Trump administration has initiated a plan to utilize Google Gemini, a leading large language model, to draft federal transportation regulations. This development, confirmed by Department of Transportation (DOT) records and internal interviews, marks a significant shift in how U.S. policy may be constructed, prioritizing unprecedented speed over traditional bureaucratic deliberation.

According to reports, the initiative aims to dramatically compress the timeline for creating new safety rules—ranging from aviation standards to pipeline oversight—from months or years down to a mere 30 days. The move aligns with a broader administration directive to integrate Artificial Intelligence into federal operations, raising both enthusiastic support for efficiency and sharp concerns regarding Transportation Safety.

The "Good Enough" Standard: Compressing Timelines

The core driver of this initiative is speed. During a strategic meeting, DOT General Counsel Gregory Zerzan reportedly emphasized a departure from perfectionism in favor of rapid output. “We don’t need the perfect rule... We want good enough,” Zerzan stated, adding that the agency intends to “flood the zone” with new regulations.

The proposed workflow envisions a radical transformation of the rulemaking lifecycle. Conventionally, drafting a Notice of Proposed Rulemaking (NPRM) involves intricate legal review, subject matter expertise, and extensive drafting. Under the new plan, agency attorneys would utilize Google Gemini to generate draft regulations in minutes. A presentation delivered to over 100 DOT staff members claimed that the AI model could handle "80% to 90%" of the drafting workload, leaving human staff to proofread and finalize the machine-generated text.

The ambitious goal is to accelerate the process so that a regulation can move from an initial concept to a complete draft ready for review by the Office of Information and Regulatory Affairs within a single month. Proponents argue this will eliminate human "choke points" in the federal bureaucracy.

Safety vs. Speed: The Internal Debate

While efficiency is the primary objective, the reliance on Automated Rulemaking for critical safety protocols has sparked significant internal debate. Transportation regulations govern high-stakes areas, including the structural integrity of aircraft, the safe operation of freight trains carrying toxic chemicals, and the maintenance of gas pipelines.

Staffers have expressed alarm at outsourcing these responsibilities to AI models known for "hallucinations"—errors where the AI confidently generates incorrect or non-existent facts. During the demonstration of the tool, a presenter reportedly dismissed concerns about the complexity of regulatory preambles, characterizing them as "word salad" that Google Gemini is well-equipped to replicate.

Mike Horton, the DOT’s former acting chief AI officer, offered a stark critique of the strategy. He compared the approach to "having a high school intern" draft federal law, warning that while the administration wants to move fast, doing so in the safety sector means "people are going to get hurt."

The Role of Google Gemini in Federal Policy

The specific tool selected for this transformation is DOT's version of Google Gemini. In demonstrations, the model was shown generating documents resembling official rulemaking notices simply by inputting topic keywords. However, observers noted that the AI-generated drafts appeared to lack the specific regulatory text required for the Code of Federal Regulations, highlighting the current gap between the model's capabilities and the legal precision required for enforceable law.

This initiative is not occurring in a vacuum. It mirrors the Department of Government Efficiency (DOGE) and Elon Musk’s push to automate federal workflows. Leaked presentations from DOGE have previously suggested using AI to automate the writing of submission documents, aiming to reduce the federal workforce's involvement in drafting.

Justin Ubert, a division chief at the Federal Transit Administration, suggested a future where human involvement recedes further. Speaking at a recent summit, he predicted that humans would eventually fall back into an oversight role, monitoring "AI-to-AI interactions" rather than actively drafting content.

Comparative Analysis: Traditional vs. AI-Assisted Rulemaking

The following table outlines the projected operational shifts between current methods and the proposed AI-integrated workflow:

Expert Reactions and Future Implications

The integration of Artificial Intelligence into the drafting of binding law challenges the traditional administrative law requirement that federal rules be built on "reasoned decision-making." Legal experts warn that while AI can generate plausible-sounding text, it lacks the ability to engage in the actual reasoning required to justify complex policy decisions.

Bridget Dooling, a professor at Ohio State University, cautioned against conflating output volume with regulatory quality. “Just because these tools can produce a lot of words doesn’t mean that those words add up to a high-quality government decision,” she noted.

As the DOT moves forward with this "point of the spear" initiative, the balance between leveraging generative AI for administrative efficiency and maintaining the rigorous standards necessary for public safety remains a critical area of observation. The outcome of this experiment could set a precedent for how the U.S. government adopts technology to write the rules that govern daily life.