Weakening Security Assumptions and Oblivious Transfer

Our work is motivated by a recent trend in cryptographic research. Protocol problems that have previously been solved subject to intractability assumptions are now being solved without these assumptions. Examples of this trend include a new completeness theorem for multiparty protocols [BGW,CCD], and a protocol for byzantine agreement using private channels [FM]. These breakthroughs illustrate both the strengths and the weaknesses of using the cryptographic model. Devising first a protocol that uses cryptographic assumptions can give powerful intuition that later allows one to create a protocol that works without assumptions. However, there is a danger that the cryptographic assumptions one uses can become inextricably bound up in the protocol. It may take years before these assumptions can be ironed out of the final protocol.