On the copyright differential protection mechanism of AI-generated content

Granting copyright protection to AI-generated content that meets the objective standards of works under the Copyright Law holds multifaceted legitimacy. It is essential to establish a differentiated protection mechanism across various dimensions, including the identification of works, rights attribution, and special regulations. Specifically: Human-AI collaborative content that possesses originality may qualify as a work. Users, upon proving their creative intellectual contribution, may acquire authorship and enjoy copyright. AI-autonomously generated content that exhibits originality may be recognized as a special category of work, with the copyright property rights vested in the AI service provider. Both aforementioned types of works should be appropriately labeled upon publication to distinguish them from purely human-authored works. AI-generated content that does not constitute a work may be treated as the fruits of property, belonging to the data assets of the service provider. If such data collections represent intellectual achievements with economic value, the service provider may acquire data intellectual property rights in accordance with the law. Relevant parties may agree on the entitlement to rights and interests of the generated content, provided such agreements do not violate the principle of fairness. Prior to the amendment of the Copyright Law, the construction and implementation of this differentiated protection mechanism rely on the Supreme People’s Court issuing judicial interpretations and administrative authorities providing administrative guidance to collaboratively advance the framework.