The Trump administration reportedly plans to use artificial intelligence (AI) to draft some federal regulations. Citing U.S. Department of Transportation (DOT) records and interviews with six officials, the report claimed that the DOT may use AI to draft some federal transportation regulations. This marks what officials describe as a broader federal effort to accelerate rulemaking through AI.The plan was presented to DOT officials during an AI demonstration in January, ProPublica reported. “Potential to revolutionize the way rules are made” Daniel Cohen, the agency’s attorney, said in a letter to colleagues. Cohen said the demonstration would be a show. “The new AI tools available to DOT rulemakers will help us do our jobs better and faster.”Citing reviewed meeting notes, the report claimed that discussions about the plan continued last week among agency leaders. Gregory Zerzan, the agency’s general counsel, said at the meeting that U.S. President Donald Trump: “We are very excited about this initiative.” Zerzan positions the DOT as a leader in a wide range of federal efforts, calling the department “ “Spearhead” and “We are the first government agency to be able to fully use AI to draft rules.”Zerzan’s comments suggested a focus on quantity over quality in regulated production. “We don’t need perfect rules for XYZ. We don’t even need very good rules for XYZ. We want enough. We’re flooding the zone.” Zerzan further added that the report cited and referred to the minutes of the meetings that were reviewed.
Why some are concerned about using AI to draft DOT regulations
Some DOT members may be concerned about these developments, according to the report. The agency’s regulations cover nearly every aspect of transportation safety, including keeping planes in the air, preventing gas pipelines from exploding, and stopping freight trains carrying toxic chemicals from derailing. Some DOT staff questioned why the federal government would delegate the creation of such an important standard to an emerging technology known for introducing errors.But proponents of the plan had a simple answer: speed. Drafting and revising complex federal regulations can take months or even years. But using the DOT version of Google Gemini, employees were able to draft the rule in minutes, or even seconds, according to two DOT employees who attended the December demo and remembered what the presenters said. Additionally, most of what is written in the preamble of DOT regulatory documents is “Word Salad” One staff member recalled the presenter saying: Google Gemini can run word salads.At last week’s meeting, Zerzan reiterated his goal of leveraging AI to speed up rulemaking. The goal is to shorten the timeline for development of transportation regulations, allowing them to go from concept to draft ready for review by the Office of Information and Regulatory Affairs in as little as 30 days, he said. This should be achievable, he said. “It should take less than 20 minutes to submit a draft rule from Gemini.”The DOT’s previously unreported plan opens a new chapter in the Trump administration’s efforts to incorporate AI into federal operations. This administration is not the first to introduce AI. Over the years, federal agencies have gradually incorporated this technology into their operations for applications such as document translation, data analysis, and classification of public comments. But the current administration has shown particular enthusiasm for the technology. President Trump issued several executive orders last year supporting AI. In April, Office of Management and Budget Director Russell Vought distributed a memo urging the federal government to accelerate its use of the technology. Three months later, the government released the following report: “AI Action Plan” Contains similar directives. But none of those documents explicitly calls for using AI to write regulations, as the DOT currently plans to do.These plans are already underway. A Department of Transportation official briefed on the matter said the agency used AI to draft a previously unpublished Federal Aviation Administration rule.Critics argued that large language models such as Gemini and ChatGPT are error-prone and lack human reasoning, and should not be entrusted with the complex and important responsibility of governance. But proponents see AI as a way to automate routine tasks and improve efficiency in the slow-moving federal bureaucracy.This optimistic outlook was evident in a windowless conference room in northern Virginia earlier this month. There, federal technology officials gathered at an AI summit to discuss the adoption of AI. “AI culture” with the government “Skill up” Allow federal employees to use this technology. Federal representatives included Justin Hubert, director of DOT’s Federal Transit Administration’s Cybersecurity and Operations Division, who participated in a panel discussion about the Department of Transportation’s plan. “Rapid implementation” A.I. many people think of humans “Choke Point” He pointed out that it slows down the AI. But Uebert predicted that eventually humans would return to a mere supervisory role and monitor things. “Interaction between AI” A presentation at DOT in December 2025 featured a similar enthusiastic perspective on the potential of AI. The presentation was attended by more than 100 DOT officials, including department heads, high-ranking attorneys, and public officials from rule-making bodies. Presenters said Gemini could handle 80% to 90% of the draft regulation work, while DOT staff could complete the rest, one attendee recalled.To demonstrate this, the presenter asked the audience about topics on which the DOT might need to issue a notice of proposed rulemaking. This public filing outlines the agency’s plans to introduce new regulations or change existing regulations. They then entered topic keywords into Gemini, which generated a document similar to a notice of proposed rulemaking. But the actual language in the Code of Federal Regulations appeared to be missing, one staff member recalled.Presenters had little concern that AI-generated regulatory documents could contain so-called hallucinations, the false text frequently generated by large-scale language models such as Gemini, three attendees noted. In any case, that’s where DOT staff steps in, he said. “His vision for the future of rulemaking at DOT was like our job was to calibrate this mechanical product. He was very excited.” one employee pointed out. Meanwhile, attendees could not clearly recall the name of the lead presenter, but said they believed it was Brian Brossos, the agency’s acting chief AI officer. The December presentation left some DOT staff with serious questions. They said rulemaking is a complex task that requires expertise not only in the subject matter but also in existing statutes, regulations, and case law. Errors or oversights in DOT regulations can lead to lawsuits, injuries, and even death within the transportation system. Some rule writers have decades of experience. But the presenter seemed to ignore all of that, attendees said. “That seems so irresponsible.” One person said he, like others, requested anonymity because he was not authorized to speak publicly about the issue.Mike Horton, DOT’s former acting chief AI officer, also criticized plans to use Gemini to write regulations, comparing it to: “We have a high school intern who helps us make the rules.” He revealed that this plan was not moving forward when he left the agency in August 2025.Houghton emphasized that transportation safety regulations are a matter of life and death, and the agency’s leaders said: “You want to run fast and break things, but running fast and breaking things means people get hurt.”Academics and researchers who monitor the use of AI in government expressed mixed opinions about the DOT plan. This technology could be beneficial and time-saving for agency rulemakers to use as research assistants with substantial oversight and transparency. However, delegating too much responsibility to AI could lead to important regulatory deficiencies and violate the requirement that federal regulations be built on rational decision-making.“Just because these tools can generate a lot of words doesn’t mean that those words will lead to high-quality government decisions. I think it’s very fascinating to figure out how to use these tools, and it makes sense to try them out. But I think it should be done with a healthy dose of skepticism.” Bridget Dooling, an administrative law professor at Ohio State University, told ProPublica.Ben Winters, director of AI and privacy at the Consumer Federation of America, said the plan is particularly worrisome given the administration’s cuts to the federal workforce last year that have seen subject matter experts leave the government. Since Trump returned to the White House, the Department of Transportation has experienced a net loss of nearly 4,000 of its 57,000 employees, including more than 100 attorneys, according to federal data.Elon Musk’s Department of Government Efficiency has been a strong advocate of AI adoption in government. In July 2025, the Washington Post reported on a leaked DOGE presentation that proposed using AI to eliminate half of all federal regulations. Part of this was done by having AI draft regulatory documents. “Writing will be automated” The presentation was read aloud. DOGE’s AI program “Automatically create drafts of all filings for attorneys to edit.”The White House did not respond to questions about whether the administration plans to use AI in rulemaking for other government agencies. Four senior technology officials in the administration said they were not aware of any such plans. About DOT “Spearhead” Two of the officials expressed doubts. “There’s a lot of pretense of, ‘We want to be seen as a leader in federal AI adoption.’ I think this is a very marketing thing.” one of them pointed out.
