GENERATIVE ARTIFICIAL INTELLIGENCE IN SOFTWARE ENGINEERING: REDEFINING PROGRAMMING PARADIGMS AND DEVELOPMENT PRACTICES

Abstract

This paper explores the role of generative artificial intelligence (GenAI) in advancing software engineering from a research and implementation perspective, as it is transforming the way software is designed, built, and maintained. We explore how large language models and code synthesis tools change the fundamental programming paradigms, moving the focus from manual coding to intent-driven specification, automated refactoring, and context-aware development. To contextualize these concepts, we present case studies that evaluate GenAI-assisted coding environments for software testing, debugging, and system design. Our quantitative analysis reveals productivity gains in prototyping and defect detection, while our qualitative findings highlight some of the challenges emerging in terms of reliability, maintainability, and developer trust. The findings indicate that GenAI is more than a tool for assistance, but a force for methodological change in software engineering, reshaping developer roles and workflows. Finally, we outline open research directions for integrating generative models with formal verification, collaborative programming, and sustainable software lifecycle practices.