There is no foolproof way, but uncertainty is something the law has dealt with for centuries, and isn't unique to this situation. The short answer is: it gets argued in court like everything else.

If cryptography introduces too many new cases of contempt, we'll need to rework the contempt system, and, in particular, introduce juries to the process (perhaps for contempt sentences exceeding a certain number of weeks).

> uncertainty is something the law has dealt with for centuries, and isn't unique to this situation. The short answer is: it gets argued in court like everything else.

Yes, precisely. This is a specific point I have seen programmer types uniquely vulnerable to tripping over.

Courts, generally, aren't like computers, where a little technicality will override common sense completely in a literalistic following of instructions. Sure there are exceptions to this rule, but for the most part a court proceeding has humans running it that live in the actual society and have some ability to factor in not just the rules but also the consequences, including unintended ones, of their actions.

"Courts, generally, aren't like computers, where a little technicality will override common sense completely in a literalistic following of instructions."

True, and in rare cases where something like that does happen, it's probably going to be a lot more amenable to sensational reporting, so more likely to be widely reported.

He isn't getting the chance to argue it in court, from what i've seen.

Have you read the filings? It sure seems like he has.

> I know for myself, even with passwords I use multiple times a day from memory, it only takes a couple of months of not using them before I cannot recall them again. This guy has been locked up for 7 months.

You have the timeline reversed. He's been locked up for 7 months because he refuses to decrypt the drives.

But still - not being able to recall a password NOW is real possibility.