China Publishes First Group Standard for AI Learning Machines, Led by Yuanli Technology
China's first group standard for AI learning machines, the Technical Requirements for Precision Learning Systems of AI Learning Machines, was officially released in November 2026. Beijing Yuanli Technology Co., Ltd., the company behind the Xiaoyuan learning machine, served as the lead drafting unit and described the document as a milestone for the standardized development of China's AI education hardware sector.
The standard arrives amid rapid growth — and growing noise — in the learning device market, where AI-powered features are often marketed with vague claims. It is the first document to define three core concepts in writing: the precision learning system, the AI learning machine, and intelligent learning guidance, giving the industry a common vocabulary and, in effect, an entry threshold.
Chapter 5 (functional requirements) and Chapter 6 (performance indicators) of the standard were proposed and largely shaped by Xiaoyuan, and center on a knowledge mastery model. The benchmarks are explicit: prediction accuracy for knowledge-point mastery must reach 90% or above, coverage of exam points in the knowledge graph must reach 95% or above, and precision in identifying a student's weak knowledge points must reach 85% or above. Learning diagnosis response time is capped at 2 seconds.
Xiaoyuan said its own products already exceed these baselines, citing its learning analytics algorithms and knowledge graph, along with large-model-driven voice interaction and personalized tutoring features. The company has not disclosed independent test data to back the claim. It has also participated in drafting China Academy of Information and Communications Technology group standards for intelligent learning agents.
The commercial backdrop reinforces the push. During the recent Double 11 shopping festival, the Xiaoyuan Xuelian S6 device surpassed RMB 100 million in sales, ranking No.1 on Douyin's electronics-education chart and No.6 across all product categories on the platform, according to figures cited by the company.
The standard also addresses safety and compliance rather than pure performance: it incorporates China's Multi-Level Protection Scheme Level 2 cybersecurity requirements, protections for minors' personal information, AI ethics provisions and anti-addiction measures. Xiaoyuan said it follows principles under which student learning data is not transferred abroad, not used commercially and not applied to student rankings.
With the standard now in force, the AI education hardware market is expected to see faster consolidation, as products without verifiable precision-learning capabilities face a clearer basis for comparison. For global edtech watchers, the move signals that China's learning device market — one of the world's largest — is moving from feature competition toward benchmark-driven quality competition.