January 2025. - Cignal selected to deliver AI safety and security training at Black Hat Las Vegas 2025.
January 2025. Cignal has been selected to deliver AI safety and security training at Black Hat 2025. This class is designed to explore AI security and safety through the lens of historic technological disasters, applying these lessons to the emerging and critical field of AI/ML security and safety. "The Black Box" refers to both an opaque nature of AI systems and "black box" recording devices used to reconstruct and analyze major disasters. This lab-focused course explores the vulnerabilities of artificial intelligence systems and the methods for mitigating their risk.
The labs expose students to a variety of testing techniques, including adversarial testing and advanced subjects like counterfactual testing with generative AI. Labs are designed to broadly introduce and describe risks across classification systems, generative models, agentic architectures, and AI systems-of-systems. Participants will get hands-on experience using a variety of open source tools for AI attack and defense, and will become familiar with techniques and approaches to communicate and mitigate safety and security risks.
Read more and sign up for the class offered in-person here.
