The most likely thing is going to be that Apple is going to be asked to allow police devices to be added to a user's iMessage account without alerting the user

The most likely thing is going to be that Apple is going to be asked to allow police devices to be added to a user's iMessage account without alerting the user (but giving them access to their messages). The real problem is that the law allows them to ask an individual to become a saboteur and it's unclear if you had a system that was explicitly resilient to such attacks (signed GPLv3 code with a threshold signing scheme with each key owned by people under different jurisdictions) whether you would be forced to dismantle such a system. I think we'll need to start rethinking threat models.