We present a demonstrator with three Pepper robots verbally debating on topics chosen by a user, verbally or in a GUI. The results highlight how LLMs may be used to argue both for and against different standpoints in debates, and how the output arguments depend on a programmed bias.