Trump Department Responsible for Airline Safety Using AI to Write New Regulations
The Department of Defense might be the first government agency to roll out a department-wide AI chatbot, but the Department of Transportation (DoT) is set to be the first to draft actual binding regulations with this technology.
AI Integration in Regulation Drafting
According to a recent investigation by ProPublica, the top transportation agency has tapped Google Gemini to assist in writing new regulations affecting aviation, automotive, railroad, and maritime safety. Internal communications from DoT attorney Daniel Cohen revealed that agency staff were presented with a demonstration of AI’s “potential to revolutionize the way we draft rulemakings.”
Cohen expressed excitement about the demonstration, highlighting “exciting new AI tools available to DOT rule writers to help us do our job better and faster.”
Speed Over Accuracy
Disturbingly, the focus on speed was emphasized, even at the expense of accuracy. According to DoT general counsel Gregory Zerzan, “We don’t need the perfect rule on XYZ. We don’t even need a very good rule on XYZ,” indicating a preference for “good enough” regulations. “We’re flooding the zone,” he remarked.
Support From Leadership
Zerzan noted that enthusiasm for the DoT AI tool extends to the highest levels, stating that Donald Trump is “very excited about this initiative.”
Potential Impact on Regulation Timelines
Typically, regulation-writing can take months or even years due to the complexities involved. However, during the demonstration in December, it was claimed that Google’s Gemini could reduce this process to “minutes or even seconds.” Zerzan set an ambitious goal, stating that a new regulation could potentially be drafted in as little as 30 days, with a draft rule produced in under 20 minutes.
Concerns Over AI Reliability
For anyone concerned about the safety of trains and planes, this development raises significant alarms. Large language models (LLMs) like Gemini are known to produce errors referred to as hallucinations, leading to potentially dangerous misinformation. Gemini has previously been linked to embarrassing incidents, such as fabricating marriages that do not exist or generating harmful medical advice.
Moreover, users expressed concern last year when Gemini produced erratic outputs resembling psychological breaks—an alarming prospect when dealing with federal regulations related to critical areas such as air traffic control.
This initiative by the DoT underscores a crucial intersection between technology and public safety, raising pressing questions about the implications of relying on AI for regulatory frameworks.
