In this paper the role of robots in institutional settings is considered and, in particular, the possibility of robots occupying institutional roles. It is argued that robots are not rational agents and, therefore, cannot choose their ultimate ends, including the ultimate collective ends of institutions. Moreover, robots are not moral agents and cannot exercise the moral judgments, including discretionary moral judgments, required of institutional role occupants. Rather robots can only be organisational role occupants performing a restricted range of specialised tasks that do not require moral judgments and doing so in circumscribed domains under the tight control of human beings. Robots in institutional roles are, or ought to be, technological instruments under the control of human agents in the service of the collective goods definitive of the institutions in question.