The Promises and Perils of using LLMs for Effective Public Services

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Moon, Erina Seh-Young, Tamura, Matthew, Zhai, Angelina, Habib, Nuzaira, Shirazi, Behnaz, Kassam, Altaf, Saxena, Devansh, Guha, Shion
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911390329995264
author Moon, Erina Seh-Young
Tamura, Matthew
Zhai, Angelina
Habib, Nuzaira
Shirazi, Behnaz
Kassam, Altaf
Saxena, Devansh
Guha, Shion
author_facet Moon, Erina Seh-Young
Tamura, Matthew
Zhai, Angelina
Habib, Nuzaira
Shirazi, Behnaz
Kassam, Altaf
Saxena, Devansh
Guha, Shion
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2601_15163
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle The Promises and Perils of using LLMs for Effective Public Services
Moon, Erina Seh-Young
Tamura, Matthew
Zhai, Angelina
Habib, Nuzaira
Shirazi, Behnaz
Kassam, Altaf
Saxena, Devansh
Guha, Shion
Human-Computer Interaction
Computers and Society
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.
title The Promises and Perils of using LLMs for Effective Public Services
topic Human-Computer Interaction
Computers and Society
url https://arxiv.org/abs/2601.15163