en.Wedoany.com Reported - On May 8, 2026, the Ministry of Industry and Information Technology, the State Administration for Market Regulation, and the Ministry of Commerce of China jointly released the series of national standards for the Intelligent Classification of Artificial Intelligence Terminals (GB/Z 177—2026). This marks the first time China has established a unified intelligent evaluation and classification system covering multiple categories of AI terminals at the national level, ending the long-standing lack of a unified benchmark for defining and grading intelligence in the industry.
The standard adopts a "2+N" architectural design. The "2" refers to Part 1: Reference Framework and Part 2: General Requirements. These two foundational standards clearly define the concept of intelligence for AI terminals, the rules for level classification, and the methods for testing and assessment, forming the common basis for intelligent evaluation across all terminal categories. The "N" refers to the development of specific specifications for different terminal categories. The first batch of standards covers seven major categories: mobile phones, computers, televisions, glasses, car cockpits, speakers, and earphones, with the development of standards for other categories to follow.
The level of terminal intelligence is clearly divided into four levels. L1 is the Response Level, where the terminal can execute preset single commands and make fixed responses, representing the starting point of intelligence. L2 is the Tool Level, where the terminal can understand user intent and provide auxiliary functions in specific scenarios. L3 is the Assistance Level, where the terminal possesses strong contextual understanding and multitasking capabilities, and can proactively offer suggestions and assistance in more complex scenarios. L4 is the Collaboration Level, where the terminal can deeply collaborate with users and other devices, possessing autonomous learning and continuous evolution capabilities—this level will be further defined and refined in subsequent standard revisions based on the level of industrial development.
The establishment of the classification system directly addresses the current practical difficulties faced by the terminal industry. Concepts such as AI phones, AI computers, and AI glasses have emerged intensively since 2024, with almost all new consumer electronics products labeled "AI." However, the industry has long lacked a unified standard for measuring the level of intelligence. Products similarly labeled as "AI phones" vary greatly in computing power, model capability, and agent execution ability. Consumers find it difficult to make horizontal comparisons in purchasing decisions, and enterprises lack an authoritative benchmark for product positioning and marketing.
The standard-setting process fully embodies the characteristics of upstream and downstream collaboration within the industrial chain. The China Academy of Information and Communications Technology, as the leading unit, jointly compiled the standards with major domestic terminal manufacturers, chip enterprises, AI model providers, research institutes, and testing and certification bodies. Multi-party participation ensured the scientific nature and industrial operability of the standards, and also allowed the testing methods and evaluation indicators to align with real product development processes. From chip computing power, model capability, and interactive experience to agent task execution, the standard's evaluation dimensions cover the complete chain of terminal intelligence.
With the release of the standards, the industry is expected to witness a series of chain effects. On the product side, terminal manufacturers can clarify product intelligence positioning based on the classification system and establish benchmarking systems during the product planning and R&D stages. On the market side, unified intelligent level labeling will help consumers clearly identify the intelligence levels of different products, reducing information asymmetry. On the regulatory side, the standard's testing methods and assessment rules also provide a technical foundation for potential future market access management or certification systems.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
