Paper Type
Complete
Abstract
Generative AI (GenAI) can accelerate project documentation and cross-functional coordination in IT projects by helping teams structure fragmented information into reviewable artifacts, yet in regulated environments these benefits depend on preserving auditability, traceability, and accountability. In this study, sustainability refers to governance resilience—the institutionalization of GenAI through repeatable, auditable routines rather than environmental sustainability. Using mixed methods in two comparable regulated insurance IT projects, we combine five semi-structured interviews, a matched quasi-experiment on two log-derived indicators, and process tracing of versioned artifacts. We identify four recurring mechanisms (S1–S4), observe directional contrasts favoring the GenAI-enabled project without making causal claims, and derive a minimal governance artifact of three checkpoints with corresponding evidence loops mapped to five risk themes.
Paper Number
1268
Recommended Citation
Yang, Hangqing and Xu, Zhengchuan, "Sustainable GenAI in Regulated IT Projects: Minimal Governance Gates for Auditability and Accountability" (2026). AMCIS 2026 Proceedings. 3.
https://aisel.aisnet.org/amcis2026/sig_itpm/sig_itpm/3
Sustainable GenAI in Regulated IT Projects: Minimal Governance Gates for Auditability and Accountability
Generative AI (GenAI) can accelerate project documentation and cross-functional coordination in IT projects by helping teams structure fragmented information into reviewable artifacts, yet in regulated environments these benefits depend on preserving auditability, traceability, and accountability. In this study, sustainability refers to governance resilience—the institutionalization of GenAI through repeatable, auditable routines rather than environmental sustainability. Using mixed methods in two comparable regulated insurance IT projects, we combine five semi-structured interviews, a matched quasi-experiment on two log-derived indicators, and process tracing of versioned artifacts. We identify four recurring mechanisms (S1–S4), observe directional contrasts favoring the GenAI-enabled project without making causal claims, and derive a minimal governance artifact of three checkpoints with corresponding evidence loops mapped to five risk themes.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
SIG ITPROJMGMT