How to prove all NP statements in zero-knowledge and a methodology of cryptographic protocol design (Extended Abstract)

Abstract

Under the assumption that encryption functions exist, we show that all languages in NP possess zero-knowledge proofs. That is, it is possible to demonstrate that a CNF formula is satisfiable without revealing any other property of the formula. In particular, without yielding neither a satisfying assignment nor weaker properties such as whether there is a satisfying assignment in which x 1 = TRUE, or whether there is a satisfying assignment in which x 1 = x 3 etc. The above result allows us to prove two fundamental theorems in the field of (two-party and multi-party) cryptographic protocols. These theorems yield automatic and efficient transformations that, given a protocol that is correct with respect to an extremely weak adversary, output a protocol correct in the most adversarial scenario. Thus, these theorems imply powerful methodologies for developing two-party and multi-party cryptographic protocols.