This study presents a bibliometric analysis of Artificial Intelligence (AI) applications in software development, focusing on literature published from 2014 to 2025. Drawing on data from Scopus and Web of Science, we use quantitative citation analysis and network visualization (via VOSviewer) to identify research trends, influential contributors, and thematic clusters in the field. Five major research themes emerge—ranging from machine learning in the software development life cycle to large language models (LLMs) in agile frameworks and convolutional neural networks (CNNs) for software testing. Results show a marked increase in AI-related publications after 2018, with strong academic interest in automation, defect prediction, and process optimization. Despite this momentum, persistent challenges remain in areas such as model interpretability, data bias, and integration into existing development workflows. We highlight research gaps and propose future directions, including the need for standardized AI integration frameworks and qualitative assessments of tool adoption in real-world settings. This analysis provides a structured foundation for future research on the evolving role of AI in software engineering.