Explore the challenges and potential pitfalls of implementing AI risk management frameworks in this 34-minute conference talk from Black Hat. Delve into a critical analysis of two prominent frameworks - NIST AI RMF and the draft EU AI Act - and uncover five key roadblocks organizations may face when operationalizing these standards. Learn from qualitative studies and experimental work that highlight the unintended consequences of well-intentioned AI risk policies. Gain valuable insights into the rapidly evolving AI Risk Management ecosystem, including the proliferation of draft standards, emerging startups, and testing methodologies. Understand the implications for compliance officers, ML engineers, and security professionals as they navigate the complex landscape of AI system evaluation and risk mitigation.
Overview
Syllabus
Risks of AI Risk Policy: Five Lessons
Taught by
Black Hat