The field of privacy-preserving machine learning is moving towards more efficient and effective solutions, with a focus on homomorphic encryption and secure multiparty computation. Recent works have addressed the susceptibility of existing cryptographic methods to side channel attacks, and proposed novel mitigation strategies. Additionally, there have been significant improvements in the efficiency and accuracy of privacy-preserving federated learning frameworks, making them more suitable for real-world deployment. The integration of secure multiparty computation with machine learning frameworks has also enabled the training and evaluation of models on combined datasets from various sources, while ensuring the privacy of sensitive information. Noteworthy papers in this area include:

• Efficient Privacy-Preserving Cross-Silo Federated Learning with Multi-Key Homomorphic Encryption, which proposes a framework that reduces computation and communication overhead while maintaining comparable classification accuracy.
• Pura: An Efficient Privacy-Preserving Solution for Face Recognition, which achieves recognition speeds up to 16 times faster than the state-of-the-art.