Writing in the October 2026 issue of Tax Research, Lin Ziying examines how the risks change once AI takes over the processing of tax-related data. The shift is not merely a change of tools: application, output and cognitive paradigms, and the human-machine relationship itself, are all being transformed. The risk field moves from a closed internal setting to interconnected channels, algorithmic bias turns from explicit to implicit, information distortion moves from error to forgery, and accountability slides from clear allocation to blurred attribution, challenging the rule-of-law order in taxation and the boundary of taxpayer rights. The author suggests using the current revision of the Tax Collection and Administration Law to create a trusted and secure space for tax data, institutionalise algorithm transparency and assessment filing, establish a human-in-the-loop model, redraw the division of responsibility in human-machine collaboration, and build end-to-end governance covering prevention, monitoring and response.
News Detail
Paradigm shift and risk response in AI-driven processing of tax-related data
Tax Research (2026, No.10) publishes an article by Lin Ziying on how AI is reshaping the processing of tax-related data. Application, output and cognitive paradigms, along with the human-machine relationship, are all shifting, and the risks shift with them: the risk field moves from closed to interconnected, algorithmic bias from explicit to implicit, information distortion from error to forgery, and accountability from clear allocation to blurred attribution. The author urges using the revision of the Tax Collection and Administration Law to build a trusted secure space for tax data, algorithm transparency and filing rules, a human-in-the-loop model, and end-to-end governance.
2026-11-06