Abstract: The breakthrough of AI-Generated Content (AIGC) technology represented by ChatGPT has triggered widespread global attention, and a large number of leading enterprises with AIGC as their core business have emerged in China and the United States. The technological innovation layout of enterprises is closely related to their competitive strength, however, existing studies are less likely to provide practical reference for the development of China's AIGC industry from the comparative study of the technological innovation layout of AIGC leading enterprises in China and the U.S. Additionally, there are significant differences in the technological innovation layout between leading AIGC enterprises in China and the U.S. Therefore, there is an urgent need for a comparative study of the technological innovation layouts of the two countries. The paper collected data on over 3000 patents from 26 Chinese enterprises and 18 the U.S. enterprises, respectively. It constructed co-occurrence networks in the field of enterprise technological innovation, and through network analysis, established indicators for the technological innovation layout, including technological innovation breadth, depth, density, and separation. Based on these four indicators, the leading enterprises were clustered using K-means clustering analysis, thereby analyzing the differences in the technological innovation layout between leading enterprises in China and the U.S. It is found that that the core technological areas of technological innovation layout in AIGC enterprises in China and the U.S. are relatively consistent. There is no significant difference in technological innovation breadth (t=-0.698, p=0.489), but the U.S. enterprises perform better in technological innovation depth (t=5.891, p=0.000) and technological innovation separation (t=-2.831, p=0.004). It is evident that the U.S. enterprises tend to adopt an "offensive" technological innovation layout strategy, while China primarily adopts a defensive strategy. Furthermore, using the K-means clustering machine learning algorithm, the paper divided the 44 leading enterprises into four categories of technological innovation layouts: technological supportive, technological specialization, technological applied, and technological pioneering layouts. An analysis of the causes and consequences of these four types of technological innovation layouts was conducted. It was found that the formation of these four types of technological innovation layouts resulted from the interaction between external market changes and internal corporate innovation strategy adjustments, aiming to enhance competitiveness. Additionally, Chinese leading AIGC enterprises are lacking in technological implementation and B2B supportiveness. The paper also provides targeted recommendations for the technological innovation development of China's AIGC industry from both enterprise and industry perspectives. At the enterprise level, Chinese enterprises should focus on the R&D of core technologies, and conduct peripheral technology research based on new technologies. They should expand the field of technological innovation on the basis of forming a protection network for core technologies. At the industry level, efforts should be made to actively establish regular communication mechanisms within the industry, promote the exchange and dissemination of knowledge and information, enhance the technological innovation capabilities of enterprises, and address the shortcomings of support-type enterprises based on establishing industrial chains. Additionally, measures should be taken to prevent the technological monopoly of large enterprises from hindering the technological implementation of small enterprises.

摘要： 以ChatGPT为代表的生成式人工智能(AIGC)技术的突破引发了全球范围内的广泛关注，中美AIGC产业在技术上存在差距，亟需对其技术创新布局进行比较研究。基于中美44家AIGC企业3000多项专利数据，论文构建企业技术创新领域共现网络，通过网络分析以及机器学习等方法比较中美企业技术创新布局差异。研究发现：中美AIGC企业在技术创新宽度方面并无明显差异，但美国企业在技术创新深度和分离度上表现较好。使用K均值聚类算法，基于技术创新宽度、深度、紧密度和分离度，可以分成四类技术创新布局：技术支撑型、技术专研型、应用型和开拓型布局，基于此发现中国企业在技术落地和B2B支撑性方面有所欠缺。论文亦从企业和产业两个层面为我国AIGC产业技术发展提供了针对性建议。