In this paper, I argue that we need a more robust account of our ability and willingness to trust social robots. I motivate my argument by demonstrating that existing accounts of trust and of trusting social robots are inadequate. I identify that it is the feature of a façade or deception inherent in our engagement with social robots that both facilitates, and is in danger of undermining, trust. Finally, I utilise the fictional dualism model of social robots to clarify that trust in social robots, unlike trust in humans, must rely on an independent judgement of product reliability.