Wu and Lidar predict that when quantum networks propagate, they will likely be subject to a new type of attack. They have introduced a new concept called quantum malware to describe such an attack. In a quantum communication network, it seems reasonable to assume that there will be people trying to interfere. In fact, quantum networks offer even more opportunities for interfering. There are more ways to attack quantum computers than standard computers. Quantum computers use a qubit’s phase information. In addition to erasing qubits, a quantum hacker could add in a phase gate that changes the phase and scrambles the outcomes of the algorithms being used.

The authors have proposed a protocol that offers a solution for protecting quantum networks from quantum malware. This solution seeks to strengthen the defenders relative to the attackers, and, in return, dramatically reduces the probability of success of an attack, under reasonable assumptions. The solution involves swapping the quantum information at random times between the network and isolated, distributed “ancillas.” It applies to arbitrary attack types, provided the protective operations are not compromised.

However, the authors propose a defense that relies on the assumption that the network on times are random and secret, and that those on times are kept to an absolute minimum, assuming this combination of assumptions leads to a small probability of infection. This scheme, the authors state, will dramatically reduce the probability of success of an attack, under these assumptions.

Quantum computing also has an issue with backing up data because quantum information cannot be copied without destroying the original. So, the authors propose a backup system in which all networked quantum computers have an ancillary register of qubits equal in size to the quantum computer’s memory, which is isolated whenever the computer is linked to the network to prevent direct infection. The authors also suggest that data have to be kept somewhere safe for as long as possible. They recommend that quantum computers spend as little time as possible with their qubits exposed to the wider quantum network that may supply data.

As far as I know, this is the first paper to talk about shielding quantum computers from a malicious attack. It is an interesting topic of research, and could lead to the development of means for protecting quantum computers.