Belfast Project: Jonathan Zittrain on Cryptography for Archives
Posted on June 11, 2014
Professor Jonathan Zittrain had a very interesting op-ed in the Globe recently on lessons of the Belfast Project case for archivists. He suggests the development and use of what he calls “time-capsule cryptography” to protect archives from the subpoena power of a court.
The idea is not just that the material should be encrypted. Encryption alone would not solve the problem, since whoever has the key to decrypt the information would himself be subject to legal compulsion and could be required to disclose the key (assuming that the person holding the key is not himself subject to criminal charges, and thus that there is no constitutional issue about self-incrimination). In time-capsule cryptography, no one has the ability to decrypt the information until a certain condition (e.g., the passage of a certain period of time) has occurred. Alternately, the key could be broken up, so that several people’s fragments of the key would be necessary to decrypt the information, and those people could agree not to reveal their key fragments until a specified condition (e.g., the death of an interviewee) had occurred.
This is a creative and useful idea, though in a case like the Belfast Project it doesn’t really solve the problem. Suppose that A. interviews B., and B. admits to killing C., and suppose that the interview is tape recorded and the tape held by D. It may be that if the interview is encrypted using a method like the methods Zittrain suggests, the tape may be worthless to prosecutors and thus a subpoena to D. may be meaningless. But A. himself is subject to a subpoena, too, and could be compelled to testify to B.’s statements to him. So in a case like the Belfast Project, Zittrain’s idea works only if we assume that A. would refuse to testify even if required by law to do so. I leave it to the reader to decide whether this is a good idea.