China has now become the country with the most anti-dumping lawsuits in the world, and the trade protection of anti-dumping measures has become a huge obstacle to the sustainable development of China’s foreign trade. In view of the current situation of anti-dumping in the United States, this study combines BP neural network to construct an anti-dumping early warning model. In order to predict the longer-term future based on the existing database, the BP neural network should be used to predict the indicators in the existing index database, and then the predicted warning indicator system is used as the input layer to warn the future police. Moreover, this study conducts research on the performance of the algorithm based on the actual case analysis. The research shows that the algorithm has certain effects and can provide theoretical reference for subsequent related research.