I'm not sure this is practical, but I very much respect that people are devoting cycles to these issues, which do deserve attention. The UK law is problematic

I'm not sure this is practical, but I very much respect that people are devoting cycles to these issues, which do deserve attention. The UK law is problematic since it asks you to prove a negative, which is impossible. This tool doesn't really help, there; any data you have could be said to be 'encrypted' (one half of the symmetric pair that unlocks the desired data). If you're ever demanded to produce under that law, you are in a very bad place. More practically speaking, I think there are other improvements in these areas that can be made, particularly in situations where you want data to self-destruct if you are placed under duress, or before you have a chance to be abused for too long (a rubber hose defense). Methods like Shamir's Method can allow you to securely divvy up a decryption key any way you like among many parties and design a system around the particular needs you have - whether that's deleting the access keys for a border crossing scenario or releasing the data in an insurance scenario. But when it comes to the guy being locked in the room being asked to prove something unprovable, it's just a screwed scenario.