数字经济时代，创新是企业生存和发展的核心，是推动经济增长的关键。企业通过创新提升竞争力和效率，促进产业结构升级。人工智能技术的发展为企业创新开辟了新空间，企业通过自动化、智能化技术提高生产效率，优化业务流程，增强市场适应性，人工智能是企业高质量发展的关键。本文以工信部2019年至2022年批准建立的国家人工智能创新应用先导区事件作为准自然实验，基于2013-2023年A股上市公司数据，选择企业创新投入强度衡量企业创新能力，采用多期双重差分模型（Staggered DID）实证检验人工智能先导区政策对企业创新的影响及其作用机制，并通过平行趋势检验、倾向得分匹配（PSM）、异质性处理效应检验、替换变量、滞后周期、提前政策时间、随机分组安慰剂等多个方式进行稳健性检验。本文通过引入融资约束、代理成本和投资效率三个中介变量研究人工智能先导区政策对企业创新影响的作用机制。本文进一步从所属行业、企业规模、存续时长和产权性质等维度研究了人工智能先导区政策对企业创新影响过程中的异质性。通过模型的回归分析，本文的主要结论如下：首先，人工智能先导区政策显著提高了先导区内企业的创新水平。其次，先导区政策通过缓解先导区内企业的融资约束、股东和管理层间的代理问题和提高先导区内企业资源配置效率从而促进企业创新。最后，人工智能先导区政策对制造业企业、大规模企业、新兴企业以及民营产权性质企业创新水平促进效果更加明显。基于理论研究和实证结论，为提升企业创新意愿与能力，促进新质生产力发展，本文提出如下建议：第一，积极推进人工智能先导区建设，并根据不同地区和行业的特点，制定差异化的政策措施。同时，建立健全多元化的融资支持体系，包括政府引导基金、风险投资、银行贷款等，为企业提供充足的资金支持。第二，加强数字基础设施建设，降低企业应用人工智能技术的成本，构建低成本、高效率的技术应用生态。第三，推动产学研深度融合，加强人才培养和引进，提高人才供给的质量和效率，为企业创新提供坚实的人才支撑。
In the era of digital economy, innovation is the core of enterprise survival and development and the key to promoting economic growth. Enterprises enhance competitiveness and efficiency through innovation and promote the upgrading of industrial structure. The development of artificial intelligence (AI) technology opens up new space for innovation, and it is the key to high-quality development of enterprises to improve productivity, optimize business processes, and enhance market adaptability through automation and intelligence.This paper takes the event of national pilot zones of artificial intelligence innovation and application approved by the Ministry of Industry and Information Technology from 2019 to 2022 as a quasi-natural experiment, based on the data of A-share listed companies from 2013-2023, selects the intensity of corporate innovation investment to measure the innovation capacity of enterprises, empirically tests the impact of the policy of pilot zones of artificial intelligence on corporate innovation and its mechanism of action by using a multi-period double-difference model (Staggered DID), and Robustness tests are conducted in several ways, including parallel trend test, propensity score matching (PSM), heterogeneity treatment effect test, substitution variable, lagged period, advance policy time, and randomized group placebo. This paper investigates the role mechanism of the impact of AI pilot zone policy on enterprise innovation by introducing three mediating variables: financing constraints, agency costs and investment efficiency. This paper further investigates the heterogeneity in the process of the impact of AI pilot zone policies on enterprise innovation from the dimensions of industry, enterprise size, duration of existence and nature of property rights.Through the regression analysis of the model, the main conclusions of this paper are as follows: first, the AI pilot zone policy significantly improves the innovation level of enterprises in the pilot zone. Second, the pilot zone policy promotes enterprise innovation by alleviating the financing constraints of enterprises in the pilot zone, the agency problem between shareholders and management, and improving the resource allocation efficiency of enterprises in the pilot zone. Finally, the AI pilot zone policy has a more obvious effect on promoting the innovation level of manufacturing enterprises, large-scale enterprises, emerging enterprises, and enterprises with private property rights.Based on the theoretical research and empirical findings, in order to enhance the enterprise's willingness and ability to innovate and promote the development of new quality productivity, this paper puts forward the following suggestions: first, actively promote the construction of AI pilot zones, and formulate differentiated policy measures according to the characteristics of different regions and industries. At the same time, establish a sound and diversified financing support system, including government guidance funds, venture capital, bank loans, etc., to provide enterprises with sufficient financial support. Second, strengthen the construction of digital infrastructure, reduce the cost of enterprises applying AI technology, and build a low-cost and high-efficiency technology application ecology. Third, promote the deep integration of industry, academia and research, strengthen the training and introduction of talents, improve the quality and efficiency of talent supply, and provide solid talent support for enterprise innovation.
