Reverse Delegated Training and Private Inference via Perfectly-Secure Quantum Homomorphic Encryption

Researchers Sergio A. Ortega and Miguel A. Martin-Delgado demonstrated the first realistic implementation of perfectly-secure quantum homomorphic encryption (QHE) for quantum machine learning. Their scheme uses efficient Clifford+T decomposition to enable two key scenarios: reverse delegated training where encrypted data trains user networks via federated aggregation, and private inference where users process encrypted data on remote quantum servers. This establishes QHE as a practical framework for multi-party quantum AI.