International Open Access Journal Platform

论人工智能作为纳税主体的证成路径

The Path to Establishing Artificial Intelligence as a Taxpayer

Progress in Social Sciences / 2025,7(8): 662-667 / 2025-08-27 10 9

View Download PDF

• Authors: 杨烨
• Information:

  中国政法大学，北京

• Keywords:
  Artificial intelligence; Taxable entity; Principle of tax equity
  人工智能; 纳税主体; 税收公平原则
• Abstract: This paper examines the jurisprudential legitimacy and institutional pathways for recognizing artificial intelligence (AI) as a taxable entity, grounded in global explorations of AI taxation systems. The study identifies three justification dilemmas: First, the theoretical conflict between the subject-object dichotomy in legal status and the fiction of electronic personality. Second, dual challenges to the principle of tax equity posed by the “benefit-received” and “ability-to-pay” criteria. Third, potential risks of double taxation. For institutional construction, a tiered recognition framework is proposed, categorizing AI into instrumental, collaborative, and autonomous entities based on autonomy levels. A behavioral autonomy + economic independence criterion is established for tax liability determination, supplemented by a digital value-added accounting model and an intelligent tax agency system. International tax rules are recommended to reformulate AI permanent establishment criteria under the OECD’s Two-Pillar Solution. 本文基于全球人工智能税制探索论证人工智能作为纳税主体的法理正当性与制度路径。研究揭示三重证成困境，即法律地位层面存在主客二分理论与电子人格拟制的博弈，税收公平原则面临受益与能力标准的双重挑战，以及可能产生的双重征税风险。在制度建构层面，本文提出分层认定模式，依自主性差异将AI划分为工具型、协作型与自治型主体，构建“行为自主性+经济独立性”的纳税要件标准，增设数字增加值核算模型与智能税收代理系统。国际税收规则需依托OECD双支柱方案，重构人工智能常设机构认定标准。
• DOI: https://doi.org/10.35534/pss.0708112
• Cite: 杨烨．论人工智能作为纳税主体的证成路径［J］．社会科学进展，2025，7（8）：662-667.