AI models have the potential to reduce years of data analysis to a few minutes, creating the possibility for governance-minded researchers and policymakers to build more effective strategies and regulations in areas with large and rapidly growing volumes of information.
In a new white paper sponsored by the Vanderbilt AI Law Lab (VAILL) and Vanderbilt Private Climate Governance Lab (PCG), Vanderbilt researchers spotlight two innovative tools they built to accelerate research into climate adaptation policy and AI regulation. Understand & Build Custom LLMs to Accelerate Governance Research starts by describing how these tools work before expanding on how users can leverage the underlying framework to build their own tools for specific use cases (environmental conservation, education policy, and individual rights protection are just a few examples).
“Artificial intelligence (AI) has rapidly become an undeniable part of the governance ecosystem,” the authors write. “Of course, there are critical questions about the governance of AI, but for many practitioners, the bigger question is how to leverage AI to accelerate progress towards effective outcomes.”
The climate adaptation policy tool allows users to query, analyze, synthesize, and compare local climate adaptation plans. Officials and researchers interested in how cities or regions with similar climates handle extreme heat can use the tool to build adaptation strategies; residents concerned about flood risk could use the tool to determine whether they should move.
“Cities and counties around the nation are preparing for climate change in very different ways to meet their diverse needs,” the authors write. “However, the extent to which these plans identify and meet those needs is unclear. We developed this tool to accelerate research and city planning to understand and improve climate adaptation initiatives.”
The AI Governance Bills Tracker helps users analyze the rapidly growing number of AI regulations enacted across the country to various ends. “Consider a policy analyst working for a tech industry association who needs to advise members operating across multiple states on compliance obligations,” the authors write. “Currently, they would have to manually search for dozens of state laws, track down relevant bills, and compare definitions and requirements across dense legislative documents. With this tool, that analyst could simply search for a component of compliance, automatically uncover the right bills, and compare definitions without having to read through hundreds of pages of legislation.”
Understand & Build Custom LLMs to Accelerate Governance Research offers a detailed explanation of the underlying frame for both of these tools, as well as instructions on how to employ the framework for other topics. It estimates the effort and expertise required to complete each step, a time frame for completion, and sample use cases.
Understand & Build Custom LLMs to Accelerate Governance Research is available on the PCG and VAILL websites. The authors are Umang Chaudhry, a Senior Data Scientist at the Vanderbilt Data Science Institute; Mark Williams, a Professor of Practice at Vanderbilt Law School and the Co-Director of VAILL; Ethan I. Thorpe, the Climate Governance Fellow at Vanderbilt Law School’s PCG Lab; J.B. Ruhl, the David Daniels Allen Distinguished Chair in Law, Director of the Program on Law and Innovation, and Co-Director of the Energy, Environment, and Land Use Program at Vanderbilt Law School; and Dr. Mariah D. Caballero MS’23 PHD’25, an Assistant Professor of Environmental Studies at Wellesley College.
