Generative Artificial Intelligence (GenAI) is rapidly transforming software engineering, particularly in agile software development. It promises unprecedented productivity gains for software developers, yet its use in agile software development is unexplored. Existing research primarily examines GenAI's short-term effects on developer productivity but lacks a holistic understanding of its long-term implications on agile teams and projects. To address this gap, my study employs system dynamics modeling approach to analyze the dynamic interactions and feedback loops introduced by GenAI in agile environments. Given the collaborative and iteration-driven nature of agile environments, using GenAI leads to inevitable unintended consequences such as tech debt and defect accumulation, which I studied through the lens of the Technical Debt framework. My dissertation employs a system dynamics (SD) approach to examine how GenAI-driven coding, testing, and documentation affect agile project outcomes such as productivity and software quality over time. Building on a validated agile SD model and guided by interviews with software development professionals and real-world project data, it incorporates how GenAI use induces reinforcing and balancing feedback loops due to rapid code generation and refinement, complexity and defect buildup, and refactoring activities. The findings indicate that while GenAI initially boosts the coding productivity of individual developers, insufficient refinement of generated code can degrade the quality of the software over time. Moreover, it is found that sufficient refinement of generated code and the use of GenAI for testing and documentation can offset these drawbacks, thereby sustaining productivity and quality over the course of the project. This dissertation provides an SD model that contributes to the literature on GenAI’s effects on software development and guides individuals and organizations to effectively incorporate GenAI in their agile software projects.