While generative artificial intelligence (AI) offers tremendous benefits in reshaping software engineering, they raise pressing ethical and legal concerns, particularly around the widespread use of large-scale training datasets, often assembled through web scraping, that may contain copyrighted software code, or proprietary content without proper consent. The opaque data usage challenges existing intellectual property laws and complicates questions of ownership, attribution, and accountability. As generative AI becomes increasingly integrated into software development practices, this accountability gap undermines legal compliance, erodes trust in AI-driven tools, and hampers broader adoption of responsible AI. This project is motivated by the need to bridge this gap by developing new tools, legal insights, and accountability models to ensure that generative AI can advance responsibly while respecting copyright, licensing norms, and user rights. This project will contribute a comprehensive framework for responsible generative AI development. The proposed research focuses on (1) analyzing licensing inconsistencies and defining AI-relevant copyright interpretations, (2) uncovering memorized copyrighted code using novel prompt engineering techniques, (3) designing watermarking-based tools for identifying and mitigating unauthorized AI-generated code, and (4) developing a practical measure of accountability for generative AI in software development. In addition, the project wil