When 2/3 of nodes are malicious and take over the subnet, how would NNS know about this immediately and take action? Does NNS slash nodes that are being malicious and have not taken over the subnet yet, and how?
The last paragraph of Shuffling node memberships of subnets: an exploratory conversation - #55 by Manu seems to be relative but not completely answer this specific question. @Manu Would you mind clarifying this?