Recently, it was reported that the State Administration for Market Regulation is firmly promoting the formulation of national standards in the field of artificial intelligence. This measure not only injects strong impetus into the development of the industry, but also lays a solid foundation for standardizing market order and ensuring technical safety. China Exportsemi will conduct further analysis and discussion on the significance of this standardization.
1. Industry status: The booming artificial intelligence industry and the need for standardization
Currently, the AI industry is growing at an astonishing rate. As of November 2024, according to Zhang Yunming, Vice Minister of the Ministry of Industry and Information Technology, at the Wuzhen Summit of the 2024 World Internet Conference, the number of core AI enterprises in China has exceeded 4,500. Among them, application-layer enterprises account for more than 80%, forming a relatively complete industrial ecological chain. These companies have performed well in the fields of intelligent voice interaction, autonomous driving, medical imaging diagnosis, and intelligent manufacturing, and have continuously expanded and deepened application scenarios, demonstrating the vigorous vitality of the artificial intelligence industry.
However, with the rapid iteration of technology and the wide penetration of applications, the need for standardization has become more and more urgent. On the one hand, the quality of products and services in the market is uneven, and there is a lack of unified standards, which affects user experience and market trust. On the other hand, enterprises face many uncertainties in the process of R&D and promotion, which hinders the scale and healthy development of the industry. In this context, the standardization work of the State Administration for Market Regulation is particularly important, providing a clear direction and a solid guarantee for the standardized development of the industry.
Figure: 4,500 AI companies witness! The State Administration for Market Regulation is accelerating the standardization of the artificial intelligence industry
2. Published standards: pointing out the direction for the development of industrial standards
In June 2024, the Ministry of Industry and Information Technology, the Cyberspace Administration of China, the National Development and Reform Commission, and the National Standards Commission jointly issued the Guidelines for the Construction of a Comprehensive Standardization System for the National Artificial Intelligence Industry (2024 Edition). The guide proposes that by 2026, China will formulate more than 50 national and industry standards, participate in the formulation of more than 20 international standards, and build a standard system that leads the high-quality development of the artificial intelligence industry.
These standards are like the "road signs" of industrial development, providing a clear direction for the company's technology research and development and product application. For example, the release of general requirements for large models helps enterprises follow unified specifications when building and optimizing large models to ensure the performance, security, and reliability of the models. The evaluation indicators and methods provide a scientific basis for evaluating the performance of large models, which is convenient for enterprises to self-evaluate and compare, and also provides a reference for users to choose suitable large model products. The introduction of these standards not only standardizes the market order, but also provides strong support for the healthy development of the industry.
3. Development of multi-field standards: all-round layout leads the future of the industry
In terms of computing power platforms, the State Administration for Market Regulation is accelerating the development of standards such as deep learning compilers, high-quality datasets, and computing scheduling and collaboration. This will optimize AI data services, improve the efficiency of computing resource utilization, and provide support for building a strong AI infrastructure. For example, by developing high-quality dataset standards, you can ensure the quality and availability of data and provide a reliable data foundation for model training.
In the field of large models, the development of general large model standards such as machine vision large models and multimodal large models, as well as application and optimization technical standards such as inference engines and retrieval enhancement are also accelerating. These standards will guide the industry to better develop, select and apply large models, and promote the maturity and popularization of large model technology. Take the multimodal large model as an example, which can fuse multiple data modalities to support more complex AI applications, and the development of relevant standards will ensure its reliability and consistency in different scenarios.
In terms of embodied intelligence and intelligent twins, the deployment of standards such as intelligent voice interaction, computer vision, and knowledge graph will lead the development of cutting-edge AI technologies. These standards not only promote the progress of technology, but also provide support for the intelligent upgrade of agents. For example, the formulation of knowledge graph standards is helpful to build a more intelligent knowledge management system and improve the knowledge processing and application capabilities of agents.
In terms of the application of artificial intelligence in the industry, promoting the development of standards such as industrial models and steel industry models will help the intelligent digital transformation of traditional manufacturing industries. Focusing on life scenarios such as medical care and home furnishing, as well as the needs of key industries such as transportation, we will carry out the development of application standards to promote the deep integration of artificial intelligence and various industries. For example, in the medical field, the formulation of relevant standards will ensure the safety and effectiveness of AI technology in medical imaging diagnosis, disease prediction, etc., and bring greater value to the medical industry.
The development of standards for security governance is also indispensable. Focusing on generative AI data labeling and optimization training, we will carry out the development of data security standards to ensure the data security of the whole process of generative AI. At the same time, strengthen the development of standards such as the classification and grading of AI application security, capability maturity assessment, and ethical governance, to ensure the safety and reliability of the use of AI technology. The formulation of these standards will escort the healthy development of the artificial intelligence industry and prevent potential risks and challenges.
4. Challenges and responses to standardization: Seek a balance between norms and innovation
While standardization is progressing steadily, there are still challenges. With the rapid development of artificial intelligence technology and the continuous emergence of new algorithms, models and applications, the formulation of standards needs to keep up with the forefront of technology, be forward-looking and adaptable. At the same time, artificial intelligence involves many fields and industries, and the interests and technical characteristics of different fields are different, and how to coordinate all parties and reach a consensus is a difficult point in the process of standard formulation. In addition, the international competition for AI standards is also becoming increasingly fierce, and China needs to actively participate in the formulation of international standards and enhance its international discourse power while promoting the formulation of its own standards.
In the face of these challenges, the State Administration for Market Regulation has adopted a series of proactive and effective countermeasures. By strengthening the coordination of cross-industry and cross-domain AI standards, and launching green channels, we will improve the efficiency of standard development and ensure that standards can follow up with industry needs in a timely manner. In addition, enterprises, universities, scientific research institutions and other parties are encouraged to participate in the formulation of standards and brainstorm to ensure the scientificity, feasibility and applicability of standards. At the same time, China also actively participates in the formulation of artificial intelligence standards by international standards organizations such as ISO and IEC, promoting Chinese standards to the international market and contributing Chinese wisdom to the global artificial intelligence industry.
5. Future prospects: Standardization promotes the high-quality development of artificial intelligence
In the era of accelerated development of the artificial intelligence industry, standardization is not only an important tool to standardize the order of the industry, but also a key starting point to enhance the competitiveness of the industry. In the future, with the continuous improvement of the standard system, Chinese's artificial intelligence industry will usher in a more healthy and orderly development.
First of all, the implementation of the standard will promote the quality of AI products and services, enhance market trust, and promote the healthy development of the industry. Enterprises can optimize their technology based on unified standards to improve product reliability, safety and user experience, thereby enhancing market competitiveness.
Second, standardization will help promote the cross-industry integration of artificial intelligence technology and promote the intelligent upgrading of traditional industries. For example, industrial large models and medical AI standards will provide technical support for industries such as intelligent manufacturing and smart healthcare, and accelerate the application of artificial intelligence in various fields.
In addition, standardization will also strengthen the ethical and security governance of AI, ensuring that the technology develops while avoiding potential risks. With the improvement of security governance standards, AI will have clearer specifications in terms of data security, privacy protection, and ethical risk prevention, laying the foundation for the sustainable development of the industry.
To sum up, the artificial intelligence standardization work promoted by the State Administration for Market Regulation not only provides a clear technical route for enterprises, but also creates a good market environment for industrial development. With the continuous improvement of the standard system, Chinese's artificial intelligence industry will occupy a more favorable position in the global competition pattern and help the intelligent transformation of the economy and society.
