Assess learners' ability to design and implement privacy-preserving machine-learning pipelines using differential privacy and secure multiparty computation. Test understanding of DP concepts (ε/δ, noise mechanisms, composition, privacy accounting), MPC protocols (secret sharing, garbled circuits, secure aggregation), practical methods (DP‑SGD, federated learning), relevant frameworks, threat models, evaluation of utility–privacy–performance trade-offs, and deployment/scalability considerations.