We develop tests for proving that someone is producing text using a particular language model by correlating the text with the order of examples used to train the model.