The Promises and Perils of using LLMs for Effective Public Services
Abstract
Governments are the primary providers of essential public services and are responsible for delivering them effectively. In high-stakes decision-making domains such as child welfare (CW), agencies must protect children without unnecessarily prolonging a family's engagement with the system. With growing optimism around AI, governments are pushing for its integration but concerns regarding feasibility and harms remain. Through collaborations with a large Canadian CW agency, we examined how LocalLLM and BERTopic models can track CW case progress. We demonstrate how the tools can potentially assist workers in opportunistically addressing gaps in their work by signaling case progress/deviations. And yet, we also show how they fail to detect case trajectories that require discretionary judgments grounded in social work training, areas where practitioners would actually want support to pre-emptively address substantive case concerns. We also provide a roadmap of future participatory directions to co-design language tools for/with the public sector.
Cite
@article{arxiv.2601.15163,
title = {The Promises and Perils of using LLMs for Effective Public Services},
author = {Erina Seh-Young Moon and Matthew Tamura and Angelina Zhai and Nuzaira Habib and Behnaz Shirazi and Altaf Kassam and Devansh Saxena and Shion Guha},
journal= {arXiv preprint arXiv:2601.15163},
year = {2026}
}