This study constructs a hybrid intelligent accounting knowledge base system integrating artificial intelligence technology, aiming to improve the decision support, risk identification and compliance management capabilities in corporate financial management. The system realizes efficient data processing, dynamic update and intelligent application of knowledge through data collection, data preprocessing, knowledge representation and storage, reasoning engine and feedback optimization module. The actual performance of the system is verified based on case analysis, and its accuracy and misjudgment in risk identification and early warning are measured by tools such as ROC curve and confusion matrix, showing that the system has a high overall accuracy rate and a low misjudgment rate. The results show that the system can effectively integrate multiple financial indicators and risk factors, provide managers with timely and reliable decision support information through rule reasoning and probabilistic reasoning, and help enterprises comply with accounting regulations in compliance management to ensure the robustness of financial management. Through the feedback optimization mechanism, the system can continuously improve the accuracy of early warning and the quality of decision-making, play a positive role in the intelligentization of accounting management, and has practicality and scalability.