Novee has introduced AI Red Teaming for LLM Applications, an innovative AI-powered pentesting agent aimed at enhancing the security of AI-driven software, launched at RSAC 2026. Traditional penetration testing struggles to keep pace with rapid developments in AI applications, leaving critical vulnerabilities unaddressed. Novee’s AI agent autonomously simulates sophisticated attacks on AI applications such as chatbots and autonomous agents, identifying vulnerabilities that conventional tools and manual testing often overlook. It customizes tests by gathering context from application documentation and APIs. Unlike traditional testing, which is limited by human resources and periodic assessments, Novee’s solution enables continuous testing integrated into CI/CD pipelines. The agent supports various LLMs, including OpenAI, and adapts its techniques based on real-world attack patterns. With $51.5 million in funding, Novee aims to redefine how organizations secure evolving AI technologies, providing a necessary defense against rapidly evolving threats in AI systems.
Source link
