This chapter reviews AI risk management frameworks and approaches that have been proposed by industry groups, governments, and other entities. To start, the chapter covers literature in human-centered AI (HCAI) and identifies key components of AI risk management frameworks. Next, an overview of industry-led AI risk management frameworks for frontier AI systems is presented, including OpenAI’s preparedness framework, Google DeepMind’s Frontier Safety Framework, Anthropic’s Responsible Scaling Policy, and xAI’s Risk Management Framework. The examination then shifts to government-initiated frameworks including NIST’s AI Risk Management Framework, NIST’s Generative AI Profile, and the EU AI Act. The chapter concludes by highlighting limitations in existing risk management frameworks and outlining opportunities for future work on AI risk management approaches. Key directions for future work include (a) improving taxonomies of risk and identifying which risks ought to be prioritized, (b) operationalizing international approaches to AI risk management, (c) developing AI security strategies for AI systems and hardware while examining tradeoffs between different security approaches, (d) advancing technical work relating to the oversight and control of advanced AI systems, and (e) improving AI risk management frameworks. AI risk management frameworks could be improved by indicating specific tests to measure risks, setting concrete risk or capability thresholds, indicating specific risk mitigation measures (as well as indicating how risk mitigations will be deemed sufficient), including commitments relating to security measures, incorporating measures to improve public transparency and accountability, and drawing best practices from other fields with rigorous risk management practices.