Bronchoscopy plays an important role in the diagnosis and treatment of pulmonary diseases which can directly observe the internal conditions of the trachea and bronchi. However, the accuracy of its examination results is greatly affected by the experience, habits, and skill levels of endoscopists, resulting in unstable and even mistaken examination results, i.e., there is a lack of standardization in bronchoscopy. Meanwhile, different pulmonary endoscopists and physicians may interpret the same lesion differently, leading to inconsistent diagnostic results and corresponding treatment. Similar issues can be found in the area of nasopharyngolaryngoscopy (NPL). Artificial intelligence–assisted bronchoscopic and NPL examination, diagnosis, and treatment has become an alternative to the traditional man-powered medical paradigm. This chapter discusses current innovations and breakthroughs in the area of respiratory endoscopic artificial intelligence. Topics such as blind spot monitoring during bronchoscopic examination, bronchoscopic lesion detection, enhancement of bronchoscopic images, quality control for NPL, categorization, and identification of malignancies at different NPL parts (such as nose, vocal cord, throat, and pharynx) are mentioned. Several deep learning models and algorithms are also studied to show how respiratory endoscopic artificial intelligence is implemented and realized.