Digital News

　　The State Council recently released the Opinions on Deepening the Implementation of the “AI Plus” Initiative (hereinafter referred to as the Opinions), outlining a comprehensive plan for the initiative across key action areas, foundational support, and organizational arrangements. It is of great significance for fully harnessing AI innovation achievements and advancing high-quality development driven by AI.

　　Three New Trends in the Field of AI

　　Currently, as large models rapidly advance in capability and expand into new application areas, AI is showing three major emerging trends: diverse forms of intelligence, higher-level capabilities, and a more systematic technological framework.

　　First, diverse forms of intelligence. From language to multimodal, spatial, and physical intelligence, large models are continuously extending their capabilities, giving rise to intelligent agents in areas such as programming, scientific research, and business services. Meanwhile, intelligent terminal devices are constantly evolving. By connecting with smartphones, vehicles, other emerging devices, and Internet applications, they are driving the emergence of new business forms.

　　Second, higher-level capabilities. As their general capabilities continue to grow, large models are also rapidly advancing their specialized expertise, allowing them to learn across sectors, address highly specialized problems, and efficiently handle complex tasks such as long-term planning and human–computer interaction.

　　Third, a more systematic technological framework. As new concepts continue to emerge, such as coordinated hardware–software optimization, a 45-degree balance law for AI capability and security, and the integration of training, evaluation, and inference, AI foundational theories are continually advancing, and AI technology has evolved into a complex systems engineering discipline.

　　China has a rich array of application scenarios. Industries such as advanced manufacturing, transportation, healthcare, and wholesale and retail have a strong digital foundation, offering robust support for the effective implementation of the “AI Plus” initiative and the sustainable development of AI technologies. The “AI Plus” initiative was included in the government work reports for both 2024 and 2025, calling for efforts in areas such as industrial empowerment, end-user applications, and the cultivation of application scenarios. It underscores the continuous advancement in key and core technology R&D, the development and utilization of data, the optimization of computing power resources, and the fostering of an open and inclusive innovation environment. The Opinions focuses on the organization and implementation of key actions for the “AI Plus” initiative across six areas: science and technology, industrial development, consumption upgrading, people’s well-being, governance capabilities, and global cooperation. It explicitly defines key indicators, such as the adoption rate of AI applications and the scale of core industries in the intelligent economy, providing a clear roadmap for the “AI Plus” initiative.

　　Turning AI from a “New Technology” into “New Infrastructure”

　　Promoting the deep integration of AI with the real economy to transform it from a “new technology” into “new infrastructure” requires efforts on two fronts. On one hand, it calls for the continuous advancement in key and core technology R&D, the enhancement of foundational support capabilities, and the establishment of a robust innovation ecosystem. On the other hand, it requires a shift from technology-driven to scenario-driven approaches, addressing real industry needs and providing effective support for the implementation of the “AI Plus” initiative through measures such as security governance and financial support.

　　First, strengthen support for computing power, data, and models to achieve autonomous development under the “AI Plus” initiative. China’s high-end computing chips still have considerable room for improvement, and computing resources remain geographically dispersed with a fragmented ecosystem. It is essential to drive deep software–hardware collaborative engineering innovation across model architectures, training and inference frameworks, and computing clusters, develop heterogeneous mixed-training technologies, build physical intelligence infrastructure, and establish a nationwide unified computing network. These efforts will enhance the coordinated allocation of computing power resources and effectively support the research and development of domestic high-performance large models. In addition, large model inference, reinforcement learning training, and high-value applications, etc., are fueling a rapid increase in demand for high-quality data. It is therefore essential to further mobilize existing resources and develop new ones, improve data copyright systems, strengthen incentives for data provision, establish sustainable data operation and service management models, and advance synthetic data technologies, thereby laying a solid foundation for the next stage of AI development.

　　Second, foster an AI innovation ecosystem to support the thriving development of the “AI Plus” initiative. China has strengths in nurturing and retaining young AI talent, and it is crucial to give promising young professionals greater freedom to explore and fully unleash their creativity. Currently, domestic open-source large models have largely caught up with international technologies, showing localized strengths in specific areas of innovation. However, there remains a certain gap between China’s domestic open-source communities and foreign ones regarding the number of developers and professional influence. Building a vibrant, open, and impactful AI open-source community and exploring new models for the inclusive and efficient application of open-source models can further transform China’s AI talent advantage into technological and industrial strength.

　　Third, improve policy and regulatory frameworks to ensure the smooth implementation of the “AI Plus” initiative. Maintaining a balance between development and security, coupled with effective governance and policy support, not only safeguards against the misuse of AI but also ensures the long-term healthy growth of the AI industry. As an emerging field, AI is experiencing rapid technological evolution and accelerated industrial deployment, while our knowledge of it continues to expand in both breadth and depth. This requires us to stay highly alert and fully engaged. On one hand, we must rigorously uphold safety standards, enhance AI security research and capacity building, and expedite the creation of a dynamic, agile, and collaborative AI governance pattern. On the other hand, we should strengthen financial and fiscal support for the AI sector, establish diversified investment mechanisms, and consistently promote technological progress and industrial development in AI.

　　Strong Overall Coordination Is Key to Efficient Implementation

　　AI has three technical features—foundational, platform-based, and general-purpose—along with significant spillover effects. Regional and industry-specific characteristics must be fully considered in AI applications, while security should be managed through a nationwide, unified approach. Therefore, strong overall coordination is essential to ensure the sustained and efficient implementation of the “AI Plus” initiative.

　　First, pooling innovation resources. It is essential to actively harness the clustering effect of innovation in emerging industries, fostering the coordinated integration and efficient sharing of key resources, including models, computing power, data, talent, application scenarios, and capital, while connecting the entire AI value chain from research and development to application, thereby supporting the long-term sustainable growth of the “AI Plus” initiative.

　　Second, carry out pilot programs in an orderly manner. Priority should be given to regions and scenarios with a solid technological foundation and strong application potential, where “AI Plus” innovation pilots should be organized. These pilots should generate a series of landmark outcomes, address key common challenges, and produce replicable models and practices, thereby providing guidance and demonstration for large-scale adoption.

　　Third, ensure security and controllability. On one hand, efforts should focus on enhancing the self-reliance of fundamental AI hardware, software, and core technologies to ensure the secure and controllable advancement of the “AI Plus” initiative. On the other hand, it is important to strengthen the publicity, interpretation, and risk guidance around AI applications, thereby fostering a positive, rational, and healthy public discourse.