UNFORTUNATELY THIS SQUARE TABLE HAS BEEN CANCELLED
Adversary emulation allows security teams to adopt a threat-informed defense approach by enabling seamless security testing based on real threats and observed tactics, techniques, and procedures (TTPs). Ultimately, the approach empowers continuous validation of security posture by proactively testing enterprise security strategies, including people, processes, and technology aspects. Similarly, an essential element of security GenAI is running red teaming engagements to detect and remediate security issues promptly.
This talk presents how to leverage adversary emulation for GenAI Red Teaming exercises. By taking advantage of the TTPs described in the MITRE ATLAS, we can ensure the safety and security of GenAI systems. This adversary emulation strategy includes aspects of security chaos engineering, allowing for means to address GenAI’s unpredictability.
