With the rapid development of big data and artificial intelligence technologies, the accounting and auditing industry is undergoing unprecedented changes. Traditional risk assessment and anomaly detection methods are unable to cope with the complex and ever-changing business needs when processing massive amounts of data and identifying potential risks. Therefore, exploring the application of artificial intelligence in accounting and auditing, especially the development of intelligent risk assessment and anomaly detection systems, has become an urgent need to improve audit efficiency and accuracy. This paper first outlines the application background and importance of artificial intelligence in the field of accounting and auditing, and points out that the martificial intelligencen challenge currently faced is how to effectively use artificial intelligence technology to achieve efficient analysis of financial data and accurate risk assessment, as well as automatic identification of abnormal behavior. Subsequently, this paper deeply analyzes the shortcomings of existing research and introduces in detartificial intelligencel the design principles of the proposed intelligent risk assessment and anomaly detection system. After that, the intelligent risk assessment and anomaly detection system is constructed and the functions of the system are experimentally tested. Through experimental verification, the system improves the accuracy of risk assessment and the time of data processing. Its lowest accuracy rate is 88%, and can accurately identify abnormal behavior in financial data and provide strong support for auditors.