An example, an admin could have a personal password that is used to update a secure hash of "last known time the market was not compromised". The users can use the public key to verify the markers. Every day or two, the admin updates that marker using the private key (password). A few days after they're arrested, they won't have updated it, the users will know something is up.
This is an over engineered solution, and I never understood the value of these warrant canaries etc.
In the right set of circumstances you would be compelled to hand over any keys (with their passphrase) in an investigation like this, and failure to do so would result in harsher penalties.
This is fanciful. It assumes both extreme competence and legal powers to compel just about anything, neither of which generally exist.
You can often be forced to unlock encrypted data as part of evidence gathering, but there's no legal power to make you conduct arbitrary processes which only exist in your head.
The founder of the Silk Road was caught because he slipped up and gave away his identity. The founder of Alphabay was caught because he put his personal email address in the welcome emails.
The police relied on opsec faulires, not technical measures to capture them.
