公共数据开放能否促进人工智能创新水平的提升？——来自政府开放数据平台上线的准自然实验

Can Public Data Openness Enhance Artificial Intelligence Innovation Levels?—A Quasi-Natural Experiment from the Launch of Government Open Data Platforms

摘要: 公共数据开放是数字经济发展的重要基础性制度安排。本文将我国城市公共数据开放平台上线视为一项准自然实验，运用多期双重差分模型，基于2009~2023年全国292个城市的面板数据，系统考察了公共数据开放对人工智能创新水平的影响及其内在作用机制。研究显示，公共数据开放显著促进了城市人工智能创新，其作用机制并非单纯的数据可得性提升，而是通过降低信息摩擦与构建制度信任，推动创新要素优化重组。异质性分析表明，政策效果在知识产权保护完善、经济基础较好的东部地区更为突出。本文从要素配置与协同创新双重视角，揭示了公共数据开放推动人工智能创新的内在路径，为构建数据驱动的高质量创新生态系统提供了理论依据与政策参考。

Abstract: The opening of public data constitutes a vital institutional provision for digital economic development. Treating the launch of China’s urban public data platforms as a quasi-natural experiment, this study employs a multi-period difference-in-differences model. Utilising panel data from 292 cities nationwide spanning 2009~2023, it systematically examines the impact of public data openness on artificial intelligence innovation levels and its underlying mechanisms. Findings reveal that public data openness significantly stimulates urban AI innovation, with its mechanism extending beyond mere data accessibility. Instead, it facilitates the optimisation and reorganisation of innovation factors by reducing information friction and fostering institutional trust. Heterogeneity analysis indicates that policy effects are more pronounced in eastern regions with robust intellectual property protection and stronger economic foundations. From dual perspectives of factor allocation and collaborative innovation, this paper reveals the intrinsic pathways through which public data openness empowers AI innovation, providing theoretical foundations and policy references for constructing data-driven high-quality innovation ecosystems.