"NJIT PhD Researcher Develops Secure Code Generation System, Achieves Early Conference Acceptance"
Security vulnerabilities are a major issue in Artificial Intelligence (AI)-powered code generation. Therefore, Khiem Ton, a Ph.D. student, and his colleagues at the New Jersey Institute of Technology (NJIT) developed "SGCode," a system that uses advanced AI and security analysis tools to detect and fix security flaws during code creation. SGCode includes Large Language Models (LLMs) such as GPT-4, a graph-based Generative Adversarial Network (gGAN), and security analysis tools. The flexible system lets users switch between code security optimization methods. It delivers comprehensive security analysis reports and performance insights. This article continues to discuss the work on secure code generation.
Submitted by grigby1
Submitted by Gregory Rigby
on