The important decision is In re Grand Jury Subpoena Duces Tecum Dated March 25, 2011. From the opinion by Judge Tjoflat:
We hold that the act of Doe’s decryption and production of the contents of the hard drives would sufficiently implicate the Fifth Amendment privilege. We reach this holding by concluding that (1) Doe’s decryption and production of the contents of the drives would be testimonial, not merely a physical act; and (2) the explicit and implicit factual communications associated with the decryption and production are not foregone conclusions.
First, the decryption and production of the hard drives would require the use of the contents of Doe’s mind and could not be fairly characterized as a physical act that would be nontestimonial in nature. We conclude that the decryption and production would be tantamount to testimony by Doe of his knowledge of the existence and location of potentially incriminating files; of his possession, control, and access to the encrypted portions of the drives; and of his capability to decrypt the files.
We are unpersuaded by the Government’s derivation of the key/combination analogy in arguing that Doe’s production of the unencrypted files would be nothing more than a physical nontestimonial transfer. The Government attempts to avoid the analogy by arguing that it does not seek the combination or the key,
but rather the contents. This argument badly misses the mark. In Fisher, where the analogy was born, and again in Hubbell, the Government never sought the “key” or the “combination” to the safe for its own sake; rather, the Government sought the files being withheld, just as the Government does here. Hubbell, 530 U.S. at 38, 120 S. Ct. at 2044 (trying to compel production of documents); Fisher v. United States, 425 U.S. at 394–95, 96 S. Ct. at 1572–73 (seeking to access contents possessed by attorneys).
Requiring Doe to use a decryption password is most certainly more akin to requiring the production of a combination because both demand the use of the contents of the mind, and the production is accompanied by the implied factual statements noted above that could prove to be incriminatory. See Hubbell, 530 U.S. at 43, 120 S. Ct. at 2047. Hence, we conclude that what the Government seeks to compel in this case, the decryption and production of the contents of the hard drives, is testimonial in character.
Moving to the second point, the question becomes whether the purported testimony was a “foregone conclusion.” We think not. Nothing in the record before us reveals that the Government knew whether any files exist or the location of those files on the hard drives; what’s more, nothing in the record illustrates that the Government knew with reasonable particularity that Doe was even capable of accessing the encrypted portions of the drives. . . .
To be fair, the Government has shown that the combined storage space of the drives could contain files that number well into the millions. And the Government has also shown that the drives are encrypted. The Government has not shown, however, that the drives actually contain any files, nor has it shown which of the estimated twenty million files the drives are capable of holding may prove useful. The Government has emphasized at every stage of the proceedings in this case that the forensic analysis showed random characters. But random characters are not files; because the TrueCrypt program displays random characters if there are files and if there is empty space, we simply do not know what, if anything, was hidden based on the facts before us. It is not enough for the Government to argue that the encrypted drives are capable of storing vast amounts of data, some of which may be incriminating. In short, the Government physically possesses the media devices, but it does not know what, if anything, is held on the
encrypted drives. Along the same lines, we are not persuaded by the suggestion that simply because the devices were encrypted necessarily means that Doe was trying to hide something. Just as a vault is capable of storing mountains of incriminating documents, that alone does not mean that it contains incriminating documents, or anything at all.
Based on a very quick skim, the analysis seems mostly right to me — in result, at least, although perhaps not as to all of the analysis. I hope to blog more on the case later on when I have a bit more time.
Also note that the court’s analysis isn’t inconsistent with Boucher and Fricosu, the two district court cases on 5th Amendment limits on decryption. In both of those prior cases, the district courts merely held on the facts of the case that the testimony was a foregone conclusion.