Over 2,500 participants rated short stories from ChatGPT higher than those written by humans. These readers failed to distinguish between the two sources, performing no better than random chance. However, scores plummeted once the machine origin was disclosed. This reveals a strong psychological bias against synthetic creativity despite the actual quality of the output.