Developers increasingly rely on AI models to generate code to boost productivity and enhance efficiency. However, there remain some quality concerns regarding the AI-generated code as the generated code is produced by models trained on publicly available code, which are known to contain bugs and quality issues. Those issues can lead to trust and maintenance challenges in the end software product and throughout the development timeline. This PhD research aims to understand the nature of bugs in AI-generated code, classify the types and patterns of these bugs, and based on this classification, investigate strategies for mitigation and fixes to reduce bug-proneness in generated code.