The Jurisprudence Reconstruction of the Criminal Law Regulation of AI Products: Based on the Perspective of Art Ontology

The iterative evolution of generative artificial intelligence technology is fundamentally reshaping the essential form of creation and expression, posing a paradigmatic challenge to the criminal legal system built on the cornerstone of human subjectivity. This article systematically analyzes the essential tension between "human creation" and "technical generation" in AI-generated content through the theoretical lens of art ontology, revealing the structural deficiencies of traditional criminal law's behavior theory, principles of responsibility, and concepts of legal interests in addressing algorithmic autonomy. Research indicates that the legal characterization dilemma of AI-generated content stems not only from the novelty of the technical object but also profoundly touches upon the philosophical presuppositions of the legal system regarding basic categories such as "creation," "behavior," and "responsibility." Through a comparative analysis of the European Union's "Artificial Intelligence Act" risk classification framework, the United States Copyright Office's pragmatic stance, Japan's "quasi-work" system innovation, and China's judicial conservative posture, this article proposes constructing a multi-layered criminal regulation system with "technical expression" as its core. This system builds a differentiated attribution model based on the degree of human creative contribution, the possibility of infringement of legal interests, and specific application scenarios, effectively safeguarding core legal values such as individual legal interests, social order, and national cultural security while ensuring technological innovation. The theoretical innovation of this research lies in transforming highly abstract art philosophical theories into an operational legal argumentation framework, providing a solid legal foundation for the transformation of the criminal law system in the age of artificial intelligence.