Postquantum eavesdropping without superluminal signaling

Abstract

In a recent Letter (Phys. Rev. Lett. 95 (2005) 010503) Barrett, Hardy and Kent (BHK) considered a very interesting question which of the fundamental laws of physics ensure security of quantum cryptographic protocols. In particular, they presented quantum key distribution protocol and claimed its security in a very general framework, which allows - what they call - postquantum eavesdropping strategies i.e. which "allows for eavesdroppers who can break the laws of quantum mechanics, as long as nothing they can do implies the possibility of superluminal signaling". In this Comment, however, we show that it is still possible to construct a strategy, which although consistent with all asumptions explicitely stated in BHK paper, enables perfect eavesdropping in their protocol.