The Trump administration is reportedly planning to use artificial intelligence (AI) to draft some federal rules. A report citing records from the US Department of Transportation (DOT) and interviews with six agency staffers claimed that DOT may use AI to draft some federal transportation regulations. This marks what officials describe as a broad federal initiative to accelerate rulemaking through AI.According to a report by ProPublica, the plan was presented to DOT staff in January during a demonstration of AI’s “potential to revolutionise the way we draft rulemakings”, agency attorney Daniel Cohen wrote to colleagues. Cohen stated the demonstration would showcase “exciting new AI tools available to DOT rule writers to help us do our job better and faster.”
The report cited reviewed meeting notes to claim that discussion of the plan continued among agency leadership last week. Gregory Zerzan, the agency’s general counsel, said at that meeting that US President Donald Trump is “very excited about this initiative.” Zerzan positioned the DOT as leading a broader federal effort, calling the department the “point of the spear” and “the first agency that is fully enabled to use AI to draft rules.”Zerzan’s comments suggested a focus on volume over quality in regulation production. “We don’t need the perfect rule on XYZ. We don’t even need a very good rule on XYZ. We want good enough. We’re flooding the zone,” Zerzan added, the report cited the reviewed meeting notes to mention.
Why some are concerned about using AI to draft DOT rules
As per the report, some DOT members may be concerned about these developments. The agency’s regulations cover nearly every aspect of transportation safety, including rules that keep airplanes flying, prevent gas pipeline explosions and stop freight trains carrying toxic chemicals from derailing. Some DOT staffers questioned why the federal government would delegate the writing of such vital standards to an emerging technology that is known for generating errors.However, the plan’s supporters offered a straightforward answer: speed. Drafting and revising complex federal regulations can require months, sometimes years. However, with DOT’s version of Google Gemini, employees could draft a proposed rule in minutes or even seconds, according to two DOT staffers who attended the December demonstration and remembered the presenter saying so. Besides, most of what appears in the preambles of DOT regulatory documents is just “word salad,” one staffer recalled the presenter saying. Google Gemini can do word salad.At last week’s meeting, Zerzan reiterated the goal of speeding up rulemaking with AI. The objective is to shorten the timeline for producing transportation regulations, allowing them to go from concept to complete draft ready for review by the Office of Information and Regulatory Affairs in just 30 days, he said. This should be achievable, he said, because “it shouldn’t take you more than 20 minutes to get a draft rule out of Gemini.”The DOT plan, which has not been previously reported, opens a new chapter in the Trump administration’s effort to incorporate AI into federal government operations. This administration is not the first to adopt AI; federal agencies have been gradually integrating the technology into their operations for years, including for document translation, data analysis and categorising public comments, among other uses. However, the current administration has shown particular enthusiasm for the technology. Trump issued several executive orders supporting AI last year. In April, Office of Management and Budget Director Russell Vought distributed a memo calling for accelerated use of the technology by the federal government. Three months later, the administration published an “AI Action Plan” containing a similar directive. However, none of those documents explicitly called for using AI to write regulations, as DOT is now planning to do.These plans are already underway. The department has employed AI to draft a still-unpublished Federal Aviation Administration rule, according to a DOT staffer briefed on the matter.Critics argued that large language models such as Gemini and ChatGPT shouldn’t be entrusted with the complex and significant responsibilities of governance, given that they are error-prone and lack human reasoning. However, advocates view AI as a means to automate routine tasks and improve efficiency within a slow-moving federal bureaucracy.This optimistic perspective was evident in a windowless conference room in Northern Virginia earlier this month, where federal technology officials gathered at an AI summit to discuss adopting an “AI culture” in government and “upskilling” the federal workforce to use the technology. The federal representatives included Justin Ubert, division chief for cybersecurity and operations at DOT’s Federal Transit Administration, who participated in a panel about the Transportation Department’s plans for “fast adoption” of AI. Many people view humans as a “choke point” that slows down AI, he noted. However, Ubert predicted that eventually, humans will step back into merely an oversight role, monitoring “AI-to-AI interactions.” A similarly enthusiastic perspective about AI’s potential characterised the presentation at DOT in December 2025, which drew more than 100 DOT employees, including division heads, high-ranking attorneys and civil servants from rulemaking offices. The presenter told them that Gemini could handle 80% to 90% of the work of drafting regulations, while DOT staffers could complete the rest, one attendee recalled.To demonstrate this, the presenter asked the audience for a topic on which DOT may need to issue a Notice of Proposed Rulemaking. This public filing outlines an agency’s plans to introduce a new regulation or modify an existing one. He then entered the topic keywords into Gemini, which generated a document that resembled a Notice of Proposed Rulemaking. However, it appeared to be missing the actual text that goes into the Code of Federal Regulations, one staffer recalled.The presenter showed little worry that the regulatory documents generated by AI could contain so-called hallucinations, which are erroneous text that is frequently generated by large language models such as Gemini, three people present noted. Regardless, that’s where DOT’s staff would step in, he said. “It seemed like his vision of the future of rulemaking at DOT is that our jobs would be to proofread this machine product. He was very excited,” one employee noted. Meanwhile, attendees could not recall the lead presenter’s name clearly, but three said they believed it was Brian Brotsos, the agency’s acting chief AI officer. The December presentation left some DOT staffers with serious doubts. They said rulemaking is complex work that requires expertise in the subject matter as well as in existing statutes, regulations and case law. Errors or oversights in DOT regulations could result in lawsuits, injuries, or even deaths in the transportation system. Some rule writers have decades of experience. However, all that seemed to be overlooked by the presenter, attendees said. “It seems wildly irresponsible,” said one, who, like the others, requested anonymity because they were not authorised to speak publicly about the matter.Mike Horton, DOT’s former acting chief AI officer, also criticised the plan to use Gemini to write regulations, comparing it to “having a high school intern that’s doing your rulemaking.” He revealed that the plan was not in the works when he left the agency in August 2025.Highlighting the life-or-death stakes of transportation safety regulations, Horton said the agency’s leaders “want to go fast and break things, but going fast and breaking things means people are going to get hurt.”Academics and researchers who monitor the use of AI in government expressed varied opinions about the DOT plan. If agency rule writers use the technology as a research assistant, with substantial supervision and transparency, it could be beneficial and save time. However, if they delegate too much responsibility to AI, that could result in deficiencies in critical regulations and violate a requirement that federal rules be built on reasoned decision-making.“Just because these tools can produce a lot of words doesn’t mean that those words add up to a high-quality government decision. It’s so tempting to figure out how to use these tools, and I think it would make sense to give it a try. But I think it should be done with a lot of scepticism,” Bridget Dooling, a professor at Ohio State University who studies administrative law, told ProPublica.Ben Winters, the AI and privacy director at the Consumer Federation of America, said the plan was especially concerning given the departure of subject-matter experts from government due to the administration’s cuts to the federal workforce last year. DOT has experienced a net loss of nearly 4,000 of its 57,000 employees since Trump returned to the White House, including more than 100 attorneys, federal data shows.Elon Musk’s Department of Government Efficiency was 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, in part by having AI draft regulatory documents. “Writing is automated,” the presentation read. DOGE’s AI program “automatically drafts all submission documents for attorneys to edit.”The White House did not answer a question about whether the administration is planning to use AI in rulemaking at other agencies as well. Four senior technology officials in the administration said they were not aware of any such plan. Regarding DOT’s “point of the spear” claim, two of those officials expressed doubt. “There’s a lot of posturing of, ‘We want to seem like a leader in federal AI adoption.’ I think it’s very much a marketing thing,” one of them noted.
About the Author
