Abstract：Open and shared AI public corpora are conducive to promoting the research and development of artificial intelligence technology and the popularization of its applications, and are regarded as a new strategic resource for shaping future digital competitiveness. At present, the construction, sharing and opening of AI public corpora in China are still in the initial exploration stage, facing many dilemmas and challenges in practice. Taking local government practices in China as research samples, this study adopts the case study method to systematically explore the development status, practical dilemmas and formation mechanisms of the construction, sharing and opening of AI public corpora in China. The findings indicate that the construction, sharing and opening of AI public corpora in China currently face four main dilemmas: don't understand, unwilling, unable, and dare not. Their respective causes are insufficient cognition and conceptual confusion, insufficient motivation and lack of incentives, capability shortcomings and collaboration barriers, and security risks and compliance concerns, forming a causal chain of cognition-motivation-capability-responsibility that ultimately produces a lock-in effect. Based on these findings, it is recommended to adopt multi-pronged measures from the aspects of conceptual cognition enhancement, incentive mechanism innovation, technical capability investment, and security compliance assurance to promote the high-quality construction, efficient sharing and orderly opening of AI public corpora.
Keywords：Artificial Intelligence; Corpus; Data Sharing; Data Opening; Authorized Operation
Author：Wang Xiang , He Dan , Zheng Lei
Source: E-Government, 2026, No. 04 (Total No. 280)
Publication Date: 2026/4/28
