【摘要】 人工智能技术的迅猛发展，尤其是生成式人工智能的突破性应用，彻底颠覆了传统内容创作模式与产业生态，也对以“人类创作中心主义”为核心逻辑的版权制度发起了全方位挑战。从版权制度的理论基础与人工智能自主发展的必然性出发，剖析人工智能时代版权制度在作品认定、作者认定及训练数据合法性等方面面临的核心困境；探讨侵权认定中思想与表达二分法的适用难题、多主体共同侵权的责任分配机制及举证责任转移的实现路径；聚焦国家技术与经济主权同版权私权保护的博弈关系，探索二者平衡的制度构建路径。更多还原

【Abstract】 The rapid advancement of artificial intelligence technologies—particularly the breakthrough applications of generative artificial intelligence—has fundamentally reshaped traditional modes of content creation and the industrial ecosystem, while posing comprehensive challenges to copyright regimes grounded in a “human-centric authorship” paradigm. From the perspective of the theoretical foundations of copyright law and the inevitability of autonomous artificial intelligence development, this article examines the core dilemmas confronting copyright systems in the artificial intelligence era, including the qualification of works, the identification of authorship, and the legality of training data. It further explores the difficulties in applying the idea–expression dichotomy in infringement determinations, the allocation of liability among multiple contributing actors, and the feasible pathways for shifting the burden of proof. Finally, the article focuses on the dynamic tension between national technological and economic sovereignty and the protection of private copyright interests, exploring institutional mechanisms to achieve an equilibrium.更多还原