From Obstacles to Strategies: Integrating Artificial Intelligence into the Software Development Life Cycle

Authors: Marcela Paladino, Claudia Pons

Journal: JAIIO, Jornadas Argentinas de Informática, Vol. 11, Issue 2, Pages 101-117, 2025

Abstract: In this study, a narrative review was conducted on the integration of AI (artificial intelligence) in the SDLC (Software Development Life Cycle), analyzing both technological advancements and the obstacles preventing full adoption. Through a systematization of information extracted from recent studies, the main AI tools applied in each phase of the SDLC were identified and classified, along with the associated challenges, which are grouped into technical, organizational, and ethical dimensions. The findings reveal that, although the application of techniques such as deep learning, natural language processing, and intelligent agents offers opportunities to optimize key processes (such as planning, design, code generation, testing, and maintenance), significant barriers persist that limit the effectiveness of these strategies. Among the challenges highlighted are data quality and structuring, compatibility with legacy systems, organizational resistance, and the lack of transparency in AI models. Consequently, comprehensive strategies are proposed that combine technical improvements, extensive training programs, and robust ethical frameworks to promote a more sustainable and efficient integration of AI.

Keywords: Ethical Dimensions, Organizational, Technical, Integra- tion Challenges, Software Development Life Cycle, Artificial Intelligence

DOI / Source: https://revistas.unlp.edu.ar/JAIIO/article/view/19539

License: Creative Commons Attribution 4.0