According to the Report on the Development of Generative AI Applications (2025) (hereinafter referred to as the Report) recently released by China Internet Network Information Center, as of June this year, the number of users of generative AI in China has reached 515 million, with a penetration rate of 36.5%. In the first half of this year alone, the number of users has increased by 266 million. In the view of Liu Yulin, director of China Internet Network Information Center, the explosive growth in the number of users of generative AI in China is a clear signal that the development of digital technology has entered a new stage, reflecting two key trends: intelligence will achieve a critical leap forward, and inclusive will go deep and implement. From the perspective of intelligence, China's artificial intelligence will move from being "usable" to being "easy to use". Artificial intelligence technology is no longer a concept in the laboratory, but has formed a mature service system that can stably support billions of users and accurately respond to diverse needs. Behind this is the dual maturity of technological and service capabilities, indicating that China's artificial intelligence industry has a solid foundation for large-scale implementation and application, and will accelerate towards the stage of "deep practicality". At present, China has formed a complete artificial intelligence industry system covering the basic layer, framework layer, model layer, and application layer, and the industry chain covers various links such as chips, computing power, data, and platforms. At the same time, the application scenarios continue to expand, and exploration and practice are becoming increasingly in-depth. As of August this year, a total of 538 generative artificial intelligence services have been registered in China, and 263 generative artificial intelligence applications or functions have been registered. They are widely used in intelligent search, content creation, office assistants, intelligent hardware and other scenarios, and have also been actively practiced in agricultural production, industrial manufacturing, scientific research and other fields. From the perspective of inclusiveness, China's artificial intelligence will move from "trial" to "commonly used". Just as the Internet broke the barrier of connection, generative AI is dismantling the high wall of technology application. Whether in first tier cities or county-level rural areas, professionals or ordinary users can enjoy the convenience and efficiency brought by technology through simple interaction. This marks the shift of the dividend of artificial intelligence technology from "exclusive to a few" to "shared by the majority", becoming a new force driving inclusive social development. Artificial intelligence, as a strategic technology leading a new round of technological revolution and industrial transformation, is profoundly changing the way humans produce and live. ”Zhang Xiao, deputy director of China Internet Network Information Center, introduced that the Report, based on the overall framework of "user popularization - industrial development - typical applications - development environment", studied and analyzed the user use, industrial application and policy formulation of generative AI. The report presents the current development status of China's generative artificial intelligence industry: users' requirements for generative artificial intelligence are constantly increasing, and their processing efficiency for complex tasks is more valued in various practical application scenarios. This significantly enhances the logical reasoning ability of generative artificial intelligence, promotes the continuous expansion of model parameter scale, and continuous optimization of architecture design. In addition, the understanding, generation, and cross modal coordination capabilities of generative artificial intelligence in processing multimodal data have also been significantly enhanced. Domestic models not only improve their generation effects in semantic response, dynamic quality, and visual aesthetics, but also understand multimodal reference information such as images and videos provided by users, thereby enhancing the model's ability to understand the author's creativity. Lightweight models have become one of the trends, and related enterprises are continuously improving the performance of their main models while actively launching lightweight models to meet the needs of users in different application scenarios. Both domestic and overseas manufacturers actively release lightweight models to provide users with low-cost and easy to deploy solutions. ”Zhang Xiao said. From the perspective of user application scenarios, the main application scenarios of generative artificial intelligence products include answering questions, daily office work, leisure and entertainment, and creating content (including text, images, audio, and video). Among them, users who use generative artificial intelligence products to answer questions are the most widespread, reaching 80.9%; The proportion of users used to generate or process text is 36%; The proportion of users used to generate images and videos is 33%. Zhang Xiao told reporters that in the next stage, the development of generative artificial intelligence is expected to achieve new results in five aspects: first, model integration will create a new form of artificial intelligence; second, open source communities will provide new impetus for technological progress; third, embodied intelligence will bring new interactive experiences to users; fourth, intelligent agents will expand new boundaries of artificial intelligence capabilities; fifth, improving governance will take artificial intelligence to a new level. (New Society)