The Strategic Evolution of Software Engineering in the Generative AI Era

Abstract

The rapid proliferation of Generative Artificial Intelligence (GenAI) across the software development life cycle (SDLC) is driving a fundamental paradigm shift in software engineering. While isolated, task-level acceleration has been demonstrated under controlled experimental conditions, enterprise environments face a pronounced “productivity paradox” characterized by legacy integration friction, codebase volume inflation, and significant verification overhead. This paper investigates the macroeconomic, architectural, legal, and workforce transformations shaping AI-augmented software development. We present the Phased Impact Model for enterprise adoption, delineate a six-layer reference architecture for the Agentic SDLC (A-SDLC) and explore the restructuring of engineering labor and computer science education. Finally, strategic recommendations are outlined to guide engineering organizations in managing systemic complexity and establishing reliable, standard-compliant development pipelines.