Since air toxic waste affects both the atmosphere and people, it is a serious problem on a worldwide scale. This effort grants a thorough conversation of the AI approaches and ML algorithms used in ecological toxic waste predicting and early-cautionary schemes. Moreover, the current effort highlights additional on AI methods used for predicting numerous main toxins in part. Further attention is specified to both AI and ML methods in foreseeing long-lasting air network infections and the estimate of weather variations and temperature surfs. The bases and effects of toxins on the atmosphere and human health were enclosed in aspect, as well as the current research status on ecological toxic waste predicting methods. Related to particular AI models, the composite model makes improved and proposals developed prediction and cautionary system precision. This study absorbed on routine estimate fault guides such as R2, RMSE, MAE, and MAPE, assessing the efficiency of several AI models.