Explore privacy-preserving machine learning techniques in this 39-minute conference talk by Patricia Thaine, CEO and PhD Candidate at Private AI, University of Toronto. Gain strategic insights into implementing privacy tools such as federated learning, homomorphic encryption, differential privacy, anonymization/pseudonymization, secure multiparty computation, and trusted execution environments. Learn how to evaluate and combine these tools to achieve specific privacy goals while considering risk factors, implementation complexity, and computational resource requirements. Discover practical examples that illustrate how to approach privacy challenges in machine learning pipelines, equipping you with the knowledge to make informed decisions when developing privacy-preserving AI systems.