PromSec: An AI Algorithm for Prompt Optimization for Secure and Functioning Code Generation Using LLM

We are a community of AI/ ML/Generative AI enthusiasts/researchers/journalists/writers who share interesting news and articles about the applications of AI. 

Machine Learning News

Researchers from the New Jersey Institute of Technology and Qatar Computing Research Institute have developed PromSec, an AI algorithm designed to optimize Large Language Model (LLM) prompts for generating secure and functional code. PromSec uses a generative adversarial graph neural network (gGAN) to identify and fix vulnerabilities, establishing an iterative feedback loop with the LLM for continuous improvement. This technique not only enhances code security but also reduces operational costs and computational overhead. PromSec's unique feature is its ability to apply optimized prompts across different programming languages and LLMs, making it a scalable and reliable solution for secure software development.