Efficient Post-Quantum Secured Blind Computation

Abstract

In the medium term, quantum computing must tackle two key challenges: fault tolerance and security. Fault tolerance will be solved with sufficiently high quality experiments on large numbers of qubits, but the scale and complexity of these devices means that a cloud-based access model is likely to dominate. How can we risk evaluating valuable computations on an untrusted server? Here we detail a verifiable circuit-based model that only requires classical communication between parties. The server is blind to the details of the computation, which is computationally secure.