Presentation: Leveraging Adversary Emulation for GenAI Red Teaming

Security researcher Kennedy Torkura has outlined a method for integrating adversary emulation into generative AI security protocols. By applying MITRE ATLAS frameworks to AWS environments, organizations can better identify vulnerabilities such as data poisoning and unauthorized model access. This approach helps engineering teams proactively secure large language models and knowledge bases by adapting traditional cloud security standards to modern AI threats.