Trump DOT Plans to Use Google Gemini AI to Write Regulations — ProPublica

Source Domain: www.propublica.org

Here is a summary of the article regarding the Trump administration’s plan to use AI for federal transportation regulations, formatted as an unordered list with between 4 and 8 key points:

AI to Draft Regulations: The Trump administration aims to use AI to assist in the creation of federal transportation regulations, as revealed by U.S. Department of Transportation (DOT) records and interviews.
Speed and Efficiency: The primary motivation behind the plan is to significantly speed up the rulemaking process, aiming to reduce the timeline for turning an idea into a complete draft by up to 90%.
Concerns About Oversight: Despite the enthusiasm of agency leaders, many DOT staff and experts express concern over the reliability of AI, noting significant risks given the critical nature of transportation regulations that affect safety.
Existing Usage and Plans: While the DOT is a pioneer in this front, AI has been gradually implemented by federal agencies for various purposes. However, this initiative marks a new focus on using AI directly for regulation drafting.
AI Demonstration: The DOT presented a demonstration to staff showcasing an AI tool (possibly based on Google Gemini) to draft regulatory texts, highlighting the capability to quickly generate “word salad” content.
Regulatory Impact: Potential regulatory documents produced by AI are acknowledged to have issues like “hallucinations” or errors and will require human proofreading, a process seen as risky by some stakeholders.
Skepticism and Criticism: Internal DOT voices criticize the plan for bypassing decades of expertise in favor of potentially flawed and much faster AI-generated drafts, warning of the consequential dangers in transportation safety.
Broader Government Use of AI: The DOT’s plan fits within the broader context of the administration’s active promotion and integration of AI in federal operations, raising questions about the widespread use of AI in regulatory contexts across other agencies.