税务稽查是税收征收管理工作的重要步骤和环节，是税务机关代表国家依法对纳税人 的纳税情况进行检查监督的一种形式，作为税收征管的最后一道防线，承担着以查促收、 以查促管、以查促廉的重要职责。近年来，数字技术和人工智能的飞速发展带动生产生活 和经济形态发生了广泛而深刻的变化，也给税务部门带来了新的机遇和挑战。我国税务稽 查数字化起步较晚，相较于税务征管的数字化改革，税务稽查数字化程度较为落后，一定 程度上限制了智慧税务的全面发展。因此，细化税务稽查信息化建设，推进税务稽查数字 化进程，构建一套囊括稽查各个层面、集合稽查各项职能的现代化智慧稽查生态系统迫在 眉睫。

本文主要采用调查问卷法、访谈法、案例分析法等方式，围绕如何提高 B 市稽查数字 化水平进行研究分析，以税收遵从理论和税务风险管理理论为理论基础，以 B 市税务稽查 系统的数字化现状为研究对象，以 B 市税务稽查系统的干部调查和访谈为基本依据，以特 大虚开增值税专用发票案件为示例，在学习和借鉴国内外学者研究的基础上，梳理了 B 市 税务稽查系统的数字化中存在的一些问题，详细研判了上述问题存在的原因，结合稽查工 作的实际需要，提出了相应的改进对策和方案，为 B 市税务稽查系统提升自身的数字化水 平提供了相应的参考路径。

通过调查研究发现，B 市税务稽查数字化程度虽然整体较高，但是仍然有较大的进步 空间，稽查干部对稽查各环节税务数字化程度的评价有高有低，对税务稽查经常使用的数 字化工具评价褒贬不一。通过对基层稽查干部的实际走访和访谈，笔者收集了税务稽查在 实际办案过程中存在的一些数字化相关的问题，经过提炼归纳，总结 B 市税务稽查数字化 存在问题如下：大数据选案模型滞后、涉税相关部门间存在信息壁垒、原始数据未经多方 比对、第三方数据取得受限、税务部门内部信息交流机制不健全、税务智能化手段仍有缺 失、数字化相关制度建设滞后、稽查数字化相关人才匮乏等。上述问题涉及到稽查选案、 检查、审理、执行等各个环节，直接导致限制了 B 市税务稽查数字化的发展，一定程度上 导致办案效率低下。

经过分析研究，本文提出 B 市稽查数字化建设的建议，包括健全稽查数字化顶层设计、 完善稽查数字化相关制度、培养稽查数字化专业人才三大方向，其中在健全稽查数字化顶 层设计上，细化为完善数据采集平台、强化数据分析平台、整合数据使用平台、增强数据 安全保护、构建智慧稽查生态系统五个方面；完善稽查数字化相关制度细化为明确立法， 使得涉税信息共享有法可依和协同治理，建立常态化联合办案机制两个方面；培养稽查数 字化专业人才细化为推动组织变革和弥补人才不足两个方面。通过从制度到人才，从理论 到实践较为全面的改进提升B市稽查数字化程度，解决本文提出的稽查数字化存在的问题， 为 B 市整个智慧税务系统的建设提供新思路和新素材。

Tax inspection is an important step and link in tax collection and management, which is a form of inspection and supervision of taxpayers' tax payment by tax authorities on behalf of the state in accordance with the law. As the last line of defense in tax collection and management, tax inspection bears the important responsibilities of promoting tax collection, management, and integrity through inspection. In recent years, the rapid development of digital technology and artificial intelligence has brought about extensive and profound changes in production, life, and economic forms, and also brought new opportunities and challenges to the tax department. China's tax inspection digitalization started relatively late, and compared with the digitalization reform of tax collection and management, the digitalization level of tax inspection is relatively backward, which to some extent restricts the comprehensive development of smart taxation. Therefore, it is urgent to refine tax inspection information construction, promote tax inspection digitalization, and build a modern smart inspection ecosystem that includes all levels and functions of inspection.

This article mainly uses methods such as questionnaire survey, interview, and case analysis to study and analyze how to improve the digitalization level of tax inspection in B city. Based on the theory of tax compliance and tax risk management, the digitalization status of B city's tax inspection system is taken as the research object, and the investigation and interview of B city's tax inspection system's cadres are taken as the basic basis. Using a major case of false VAT invoices, this article summarizes some of the problems in the digitalization of B city's tax inspection system, analyzes the reasons for these problems, and proposes corresponding improvement measures and solutions based on the actual needs of inspection work, providing a reference path for B city's tax inspection system to improve its digitalization level.

Through investigation and research, it was found that although the digitalization level of tax inspection in B city is generally high, there is still room for improvement. The evaluation of tax inspection cadres on the digitalization level of tax inspection in various links varies, and the evaluation of digital tools frequently used in tax inspection is mixed. Through actual visits and interviews with grassroots inspection cadres, the author collected some digital-related problems in tax inspection in the actual case handling process. After refining and summarizing, the problems of digitalization in B city's tax inspection system are as follows: the lag of big data case selection model, information barriers between tax-related departments, original data not being compared by multiple parties, restricted access to third-party data, incomplete internal information exchange mechanism of the tax department, lack of tax intelligence means, lagging digital-related institutional construction, and shortage of digital inspection-related talents. These problems involve various links such as inspection selection, inspection, trial, and execution, which directly limit the development of digitalization in B city's tax inspection and to some extent lead to low case handling efficiency.

After analysis and research, this article proposes suggestions for the digitalization construction of tax inspection in B city, including refining the top-level design of tax inspection digitalization, improving relevant digitalization systems, and cultivating professional talents in tax inspection digitalization. Among them, the refinement of the top-level design of tax inspection digitalization includes improving the data collection platform, strengthening the data analysis platform, integrating the data usage platform, enhancing data security protection, and building a smart inspection ecosystem. The improvement of relevant digitalization systems includes clarifying legislation to provide a legal basis for sharing tax-related information and establishing a normalized joint case handling mechanism. The cultivation of professional talents in tax inspection digitalization includes promoting organizational change and making up for talent shortages. By comprehensively improving the digitalization level of tax inspection in B city from the aspects of system, talent, theory, and practice, and solving the problems of digitalization in tax inspection proposed in this article, new ideas and materials are provided for the construction of B city's entire smart taxation system.