A Norwegian team claims it has come up with a new way to hack quantum key distribution (QKD) systems that would allow an attacker to intercept a key without being detected.

An undetectable attack on QKD sounds impossible because the physics that underpin the technology are unshakable.  Anyone intercepting the photons used to encode the key in any of QKD’s sending protocols between point A and point B will alter their quantum state, making their presence immediately obvious to the receiver.

That is what, in essence, QKD amounts to – the detection of the interception of encryption keys with absolute certainty.

According to a newly published paper in Nature Photonics, however, the team at Norwegian University of Science and Technology in Trondheim decided to exploit weakness in the way QKD systems are configured by two vendors, ID Quantique and MagiQ Technologies, to break this apparently watertight assumption.

By firing a 1-milliwatt laser at the receiving photon detector, the scientists were able to ‘blind’ it, jilting it out of its quantum state. The hack was that the receiving detector has no way or realising that it is no longer operating as a quantum detector and so the unshakable laws are never brought into play.

The attack could in theory be counteracted by a separate detection system not currently fitted to early commercial QKD systems.


Normally, detection of interaction with the quantum state of an arriving photon relates to an error rate exceeding 20 percent, which takes account of the noise threshold generated by electronic devices used in QKD systems.

An attack from earlier this year was able to manipulate such noise levels to make interception below that level just possible, the first time the assumptions of QKD had been shown to be flawed in a commercial system.

The new attack uses a different method to show that QKD systems do need to be designed to take account of these theoretical attacks.

QKD creeps forward towards mainstream use in small but sometimes important advances such as one from the Toshiba Research Lab and the Cavendish Laboratory, which together found a way to use common LEDs to create photons. Distances and bitrates are also improving.