A Practical Examination of AI-Generated Text Detectors for Large Language Models