Chat-Up: The role of competition in street-level bureaucrats' willingness to break technological rules and use generative pre-trained transformers (GPTs)

• 成果类型: Article
• 署名作者: Frisch-Aviram, Neomi; Spanghero Lotta, Gabriela; Jordao de Carvalho, Luciana
• 署名单位: Getulio Vargas Foundation
• 刊物名称: PUBLIC ADMINISTRATION REVIEW
• ISSN/ISSBN: 0033-3352
• DOI: 10.1111/puar.13824
• 发表日期: 2025
• 页码: 468-485
• 关键词: artificial-intelligence; DISCRETION; Clients; implementation; ORGANIZATIONS; state

摘要

Organizations worldwide are concerned about workers using generative pretrained transformers (GPTs), which can generate human-like text in seconds at work. These organizations are setting rules on how and when to use GPTs. This article focuses on street-level bureaucrats' (SLBs) intentions to use GPTs even if their public organization does not allow its use (tech rule-breaking). Based on a mixed-methods exploratory design, using focus groups (N=14) and a survey experiment (N=279), we demonstrate that SLBs intend to break the rules and use GPTs when their competitors from the private sector have access to artificial intelligence (AI) tools. We discuss these findings in the context of hybrid forms of public management and the Promethean moment of GPTs.