Do Language Models Plagiarize?