Research on the Attribution and Liability Path of Artificial Intelligence Entities Infringing on Legal Interests

Abstract

The deep application of artificial intelligence (AI) technology has gradually permeated all aspects of social life, moving beyond a mere technological tool. The infringement of legal interests caused by AI entities is becoming increasingly prominent, posing a crucial issue for criminal law to address. This paper focuses on the attribution and liability for infringements of legal interests by AI entities. It defines the legal connotation of AI entities, distinguishes the legal attributes of weak and strong AI, categorizes the forms of infringement, and summarizes their core characteristics. Based on this, it analyzes the current challenges in criminal liability attribution, including disputes over the determination of the subject's qualifications, difficulties in determining causality, obstacles to the application of traditional attribution principles, and lagging criminal legislation. To address these issues, this paper proposes constructing a hierarchical and progressive system of responsible parties, clarifying the responsibilities of developers, users, and regulators; establishing a causal proof mechanism centered on algorithm tracing, and improving the rules for liability sharing in cases of multiple causes leading to a single effect; and constructing an internal governance path through optimized criminal legislation and improved judicial application, combined with the prioritization of civil liability and supplementary insurance and fund systems to build a multi-governance system. The aim is to provide theoretical reference and practical pathways for the criminal regulation of infringements of legal interests by AI entities, achieving a balance between the development of AI technology and the protection of criminal law.