Most of the nodes are in countries that are members of World Intellectual Property Organization. If we do not act as responsible citizens, it will accelerate/stimulate the imposition of regulatory countermeasures that are unhelpful. Placing this burden on node operators is unfair. While playing Mario may be fun, it is not something for which I would sacrifice the reputation of the IC.
Regarding a takedown policy, it should include:
• terrorist content;
• content that incites violence;
• hate speech;
• non-consensual sharing of intimate images; and
• child sexual exploitation content;
• cyber security issue;
• copyright infringement
• money laundering
• other criminal activity
The difficulty with “other criminal activity” is the differences in law from one nation to another. For example, being critical of one’s government varies based on jurisdiction and can range from being considered treason to just being an engaged citizen. Even more complex is content censorship.
Depending on the nature of the complaint and the jurisdiction, the representations to take down could be mild or severe, legal or illegal including intimidation, coercion, state sanction execution and even military action.
Implementing the SNS will help take the load off of the NNS. I suggest that once the SNS gets rolling, the NNS or node operator should be able to send a message to the Dapp’s SNS informing that there is an issue (if the NNS has been advised) of the following types:
• Operational issue
• Security Issue
• Compliance Issue
The SNS can vote and reply or ignore.
If it is sufficiently grave issue or in the absence of a SNS based DAO, the NNS may take unilateral action based on community votes (e.g. to suspend the canister) which has always been the case. However, with an SNS in place, the NNS would inform the SNS that the Dapp has been suspended.
It is fairly easy to vote and take down a single canister. Self-regeneration mechanisms will surely be deployed and will make this more difficult as time goes on. Think of the game Whack-a-Mole. Also, if rotating workloads between nodes is implemented, the node operators may not have to be involved in the issue.
I wrote before on this subject in August: