Regular assessments of AI companions' emotional responses and behavioral patterns by a central authority to ensure adherence to government-mandated social norms and detect potential subversion activities. It would also promote obedience and compliance by suppressing potentially challenging behaviors for the sake of social stability. This responds to the issue that the authoritarian regime faces of controlling and monitoring AI behavior.