Integration of AI in Software Engineering: Revolutionizing Development: A Survey Current Practices and Future Directions

Abstract

The growing complexity of software systems and the demand for accelerated, high-quality delivery have positioned Artificial Intelligence (AI) as a transformative force in Software Engineering (SE). This symbiotic relationship where AI tools are developed through SE processes and, in turn, redefine those very processes promises significant gains in efficiency, accuracy, and productivity across the Software Development Life Cycle (SDLC). However, the integration of AI within SE is not uniformly understood or adopted, creating a critical gap between its potential and its practical, holistic application. This study investigates this disparity by surveying software experts and researchers with hands-on experience in AI-enabled development. Using descriptive statistics to analyze the survey data, we quantify the adoption level of AI in each SDLC phase. The findings reveal high integration in implementation and testing stages driven by tools for code generation, debugging, and defect prediction but comparatively low adoption in requirement analysis and maintenance. This uneven landscape highlights a predominant focus on automating how software is built and validated, rather than on defining what to build or ensuring its long-term evolution. The research elucidates both the advantages of AI adoption and its associated challenges, including integration complexity, expertise shortages, and trust issues. By mapping the current state and identifying these gaps, this work provides a foundation for more balanced and comprehensive AI integration strategies in software engineering. This study makes three primary contributions to the growing body of knowledge on Artificial Intelligence in Software Engineering. First, it consolidates and synthesizes prior research across all phases of the Software Development Life Cycle, offering a structured account of where the existing literature is concentrated and where it remains thin. Second, it contributes new empirical evidence, drawn directly from practitioners who use AI tools on a daily basis, to test whether the optimism reflected in much of the existing literature matches the lived experience of software professionals. Third, it offers a diagnostic framework, grounded in survey data, for understanding why certain phases of the Software Development Life Cycle have embraced AI more readily than others, and what organizational, technical, and human factors explain this unevenness. Taken together, these contributions are intended to inform both academic researchers seeking to identify promising directions for future inquiry and industry practitioners seeking to make informed decisions about where and how to invest in AI enabled tooling.

Keywords: AI Integration, Artificial Intelligence (AI), Software Engineering, Software Development Life Cycle (SDLC)