Paper Type
ERF
Abstract
Autonomous AI agents powered by large language models are reshaping software development by transforming developer roles, team structures, and lifecycle workflows. However, systematic evidence of their operation across the full software development lifecycle (SDLC) and their governance in practice remains limited. This study presents a systematic literature review of AI agent applications in software engineering based on peer-reviewed studies from 2015 to 2026. Findings show increasing automation in code generation, testing, and deployment, along with emerging multi-agent role-based collaboration. In contrast, early SDLC phases such as requirements engineering and system design remain underexplored. Persistent challenges in trust, governance, and human-AI collaboration further limit adoption. This review synthesizes current evidence, identifies structural gaps, and proposed the need for lifecycle-integrated, human-centered governance frameworks for responsible AI agent integration.
Paper Number
1620
Recommended Citation
SERU, SAI NEELIMA and Noteboom, Cherie Bakker, "Agentic Software Engineering: A Review of AI Agents, Lifecycle Integration, and Human-Centered Governance" (2026). AMCIS 2026 Proceedings. 22.
https://aisel.aisnet.org/amcis2026/conftheme/conftheme/22
Agentic Software Engineering: A Review of AI Agents, Lifecycle Integration, and Human-Centered Governance
Autonomous AI agents powered by large language models are reshaping software development by transforming developer roles, team structures, and lifecycle workflows. However, systematic evidence of their operation across the full software development lifecycle (SDLC) and their governance in practice remains limited. This study presents a systematic literature review of AI agent applications in software engineering based on peer-reviewed studies from 2015 to 2026. Findings show increasing automation in code generation, testing, and deployment, along with emerging multi-agent role-based collaboration. In contrast, early SDLC phases such as requirements engineering and system design remain underexplored. Persistent challenges in trust, governance, and human-AI collaboration further limit adoption. This review synthesizes current evidence, identifies structural gaps, and proposed the need for lifecycle-integrated, human-centered governance frameworks for responsible AI agent integration.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
NEXTTRANS