With the rapid development of internet finance and the rapid growth of bank credit market share, the data has become complicated and diverse. The traditional risk control model, whether driven by nature or artificially induced, is incapable of meeting the current demand for bank credit default risk prediction, resulting in the occurrence of numerous default events and losses for many banks. The goal of this research is to use artificial intelligence to analyze users' credit and determine whether they will be able to repay in the future. This not only reduces the bank's lending risk, but also improves the bank's service efficiency.