This research introduces RobGen, a novel framework for systematically improving the robustness of code generated by large language models (LLMs).

• Study reveals 43.1% of LLM-generated code lacks proper robustness features
• Framework addresses input validation, error handling, and other security vulnerabilities
• Evaluation shows RobGen significantly improves code quality across multiple leading LLMs
• Provides practical solutions to make AI-generated code production-ready

Security Impact: As organizations increasingly rely on LLM-assisted programming, this framework offers critical safeguards against potential security vulnerabilities and system failures in automatically generated code.

Enhancing the Robustness of LLM-Generated Code: Empirical Study and Framework