Sure, the multiagent problem needs communication of the credible threat to the bad actors. "Credible threat" implies that bad actors don't collect information a

Sure, the multiagent problem needs communication of the credible threat to the bad actors. "Credible threat" implies that bad actors don't collect information about the system, technical and non-technical, that implies the balance of risk and reward isn't tilted in favor of continuing their actions.