Here we see the DIARC architecture controlling a Nao robot in a simple natural language based interaction. The robot is able to obey simple literal commands. It is also able to appropriately reject commands based on reasoning regarding action effects. In this case, it rejects commands that can possibly result in harm to itself.