Partners save 15% on every standard

Valid

GB/T 47746-2026Customer contact services - Requirements for collaboration between live agents and intelligent customer service (English PDF)

顾客联络服务　人工与智能客户服务协同要求

Open the GB/T 47746-2026 preview as PDF

Preview — first pages of GB/T 47746-2026 (full document: 24 pages)

This is a limited preview

Buy now to download the full PDF (24 pages)

Issued by

SAMR; SAC

Level / Type

National · Recommended

Issue date

May 25, 2026

Implementation date

September 1, 2026

Scope

GB/T 47746-2026 is the English-translated version of 顾客联络服务　人工与智能客户服务协同要求.

GB/T 47746-2026 is the Chinese national standard covering how a bot and a human agent hand work to each other in a contact centre - when the machine must escalate, what context travels with the customer, how the customer is told which they are talking to and how a request that has failed once is not sent back round the same loop. First edition, in force since 1 September 2026. It was issued on 25 May 2026 and has been in force since 1 September 2026, as a first edition. The document is under the responsibility of the Standardization Administration of China. This page is published from the official record of the 2026 edition; the clause text of a standard this recent is not yet in circulation, and the figures, limits and tables it contains are those of the document itself, delivered in full with the English translation.

Document preview — GB/T 47746-2026

National Standard of the People's Republic of China

ICS: 03.080
Classification: A 12

Issued by: State Administration for Market Regulation; Standardization Administration of the PRC

4 General Requirements
5 Inbound Call Tasks
5.1 Incoming Call Task Allocation
5.2 Inbound Customer Service Switching
5.2.2 Switching from Intelligent Customer Service to Human Customer Service
5.2.3 Switching from human customer service to intelligent customer service
6 Call up the task
6.1 Outbound Task Assignment
6.2 Switching between outbound customer service calls
7 Service Completion and Follow-up Processing
7.1 End the task
8 Information Security
9 Collaborative Service Improvements

Foreword

This document complies with the provisions of GB/T 1.1-2020 "Standardization Work Guidelines Part

1.Structure and Drafting Rules of Standardization Documents". Drafting. Please note that some content in this document may involve patents. The issuing organization of this document assumes no responsibility for identifying patents. This document was proposed and is under the jurisdiction of the National Technical Committee on Service Standardization (SAC/TC264). This document was drafted by: China National Institute of Standardization, China Southern Power Grid User Ecosystem Operation Co., Ltd., and China Academy of Information and Communications Technology. China Academy of Civil Aviation Science and Technology, Shenzhen Shifang Ronghai Technology Co., Ltd., Beijing Dongfang Caibo Enterprise Management Consulting Co., Ltd., Ant Zone Blockchain Technology (Shanghai) Co., Ltd., Beijing Branch of China United Network Communications Co., Ltd., Guangdong Branch of China United Network Communications Co., Ltd. Provincial Branch, China Mobile Online Marketing Service Center, China Unicom Online Information Technology Co., Ltd., China Telecom Consulting & Design Institute Co., Ltd., and Guangdong Dongguangxin Communication Service Co., Ltd., Guangdong United Electronic Service Co., Ltd., Hunan Provincial Expressway Network Toll Collection Management Co., Ltd. Great Wall Life Insurance Co., Ltd., Hisense Group Co., Ltd., Gree Electric Appliances, Inc., and TCL Industrial Holdings Co., Ltd. The company, Yadea Technology Group Co., Ltd., Bairong Cloud Innovation Technology Co., Ltd., Taiying Technology Group Co., Ltd., Ningbo Fotile Marketing Limited Liability Company, Qingdao Haier Lexin Cloud Technology Co., Ltd., Ninebot Technology Co., Ltd., Sichuan Tianyi Call Technology Co., Ltd., Ningbo Meimeijia Yuan Electric Appliance Service Co., Ltd., Inner Mongolia Mengniu Dairy (Group) Co., Ltd., Ningbo Qitaotao Information Service Co., Ltd., Customer Observation (Hangzhou) Digital Media Co., Ltd., Hunan Chensheng Information Technology Co., Ltd., and China Resources Gas Holdings Co., Ltd. The main drafters of this document are. Li Zipeng, Fu Qiang, Zheng Juan'er, Jin Zongzhen, Wang Qi, Kang Feng, Xu Rui, Cai Jitao, Hou Fei, Wang Nana, and Shi Zhongjin. Diao Zipeng, Yu Dingyao, Li Hongtao, Cai Huali, Wu Yansong, Lei Langsheng, Pang Junying, Liu Bin, Wang Long, Xu Ning, Fu Haiyan, Li Fangjian, Yang Nan Li Xiaoyan, Yu Lingyang, Li Shenghui, Wang Wenmin, Han Yong, Fu Gang, Liu Deming, Zhou Hang, Gong Boxuan, Gong Xiao, Shao Zefeng, Yan Li, Yu Xin, He Yilin Zhang Rongsheng, Wu Jingyu, Hong Kai, Ji Wenyao, Chen Peng, Wang Debao, Zheng Ying, Li Bianfang, Tang Linli, Fu Bin, Cheng Cheng, Lin Lei, Xiao Zijing, Cui Ming.

This document closely follows the new trend in customer contact services where intelligent customer service and human customer service are deeply integrated, focusing on solving customer issues in this scenario. The core issue is balancing customer service efficiency and service experience. This is addressed through standardized systems for inbound and outbound call processes, service termination, follow-up handling, and information processing. By improving key aspects such as security and collaborative services, we can effectively enhance customer service operational efficiency while further ensuring and optimizing the customer service experience. This document aims to achieve a dual improvement in service quality and operational efficiency for organizations across various sectors in the context of technological change. Quality development provides a comprehensive and practical standardization basis. Customer contact service Requirements for Collaboration between Human and Intelligent Customer Service

1.Scope This document specifies the general requirements, inbound call tasks, outbound call tasks, and service termination procedures for real-time collaborative work between human and AI customer service. Improvements to follow-up processing, information security, and collaborative services. This document applies to the operation and management of organizations that provide both human and intelligent customer service.

4 General Requirements

4.1 Well-established organizational support Customer contact centers should provide appropriate organizational support for the collaboration between human and intelligent customer service, including but not limited to.

---Based on industry characteristics, provide human customer service that matches user needs;

---Establish an organizational system that adapts to the collaboration between human and intelligent customer service;

---Establish emergency response mechanisms for scenarios such as disasters and system failures;

---Clearly define the organizational structure for collaboration between human and AI customer service;

---Build an internal system that includes customer information management and work order processing;

---Provide customer service access in a clear, prominent, and unambiguous manner, such as prominently displaying "Customer Service" in large print at the top of the website homepage;

---Establish a knowledge base;

---Regularly conduct intelligent customer service optimization training to train customer service personnel;

---Other organizational safeguards that support collaboration between human and intelligent customer service.

4.2 Clear division of responsibilities The responsibilities of human customer service and AI customer service should be clearly defined. AI customer service should prioritize standardization. For questions with high difficulty and clearly defined answers in the knowledge base, human customer service should prioritize those requiring multi-step analysis or integration of multiple information sources. The questions answered, and the issues that intelligent customer service failed to resolve.

4.3 Smooth service experience Customer contact centers should prioritize customer experience, seamlessly integrating human and intelligent customer service to ensure customers have a positive experience. The transition between human and intelligent customer service should be seamless, and the system should synchronously transmit customer information when switching between the two. Key data such as information and historical interaction records are collected to ensure service continuity.

4.4 Service security is controllable Customer contact centers should establish information and data security mechanisms, including data acquisition/processing/use, as well as security management of models, personnel, etc. This mechanism ensures data security and traceability throughout the entire service process, preventing data leaks and making risks traceable.

5.1 Incoming Call Task Allocation

5.1.1 General Requirements The customer service system should provide a clear, prominent, and unambiguous entry point for customer service, such as prominently displaying "Customer Service" in large print at the top of the website homepage. "Service". At the beginning of customer service, the customer service system should assess the system's problem-solving capabilities, overall call volume and handling capacity, and the customer group. (For example, the acceptance level of elderly customers) towards intelligent customer service should be considered when allocating work orders appropriately to either intelligent or human customer service for processing, and through... Provide users with the number of people in the queue or the estimated wait time via text or voice.

5.1.2 Intelligent Customer Service Processing Scenarios The customer service system should prioritize assigning tasks to intelligent customer service processing when one of the following conditions is met.

---The customer's inquiry was either a simple, repetitive, or standardized question;

---The intelligent customer service knowledge base provides clear answers to customer inquiries, enabling it to offer solutions quickly and accurately;

---In special circumstances, the inbound call volume is large, and human customer service cannot handle it immediately.

5.1.3 Human Customer Service Processing Scenarios The customer service system should prioritize assigning tasks to human customer service when one of the following conditions is met.

---The customer explicitly requested to switch to human assistance;

---Customer inquiries are often complex, including personalized needs, vague expressions, and issues involving cross-departmental coordination;

---The customer's inquiry is infrequent, for example, the question has occurred no more than 3 times in the past year;

---The knowledge base for intelligent customer service does not provide clear answers to customer inquiries;

---Scenarios involving the personal safety and property of customers;

---The customer explicitly requested to be transferred to a human customer service representative for further processing;

---The customer service system records that this customer has visited multiple times (e.g., 3 or more times within 24 hours) and has shown significant interest in intelligent customer service. dissatisfied;

---Customers in specific age groups, such as the elderly and minors, have a lower acceptance of intelligent customer service;

5.2 Inbound Customer Service Switching

5.2.1 Switching Requirements When switching between human customer service and intelligent customer service, the system should completely and accurately transmit service-related information synchronously. Key data, including customer identification information, historical interaction records, problem descriptions, and work order status, is collected to ensure the complete inheritance and continuity of service. Sexuality. During the customer wait period before the switchover is complete, the estimated wait time should be communicated to the customer via voice or text. After the switchover is complete, the customer contact... Network centers should not ask customers to repeat information they provided or questions they raised before the switch.

5.2.2 Switching from Intelligent Customer Service to Human Customer Service

5.2.2.1 Switching Method During the service process of intelligent customer service, a human service entry point should be set up to support customers switching from intelligent customer service to human customer service. The switching method should have one or more of the following features.

---Customer-initiated transfer to human operator. keyword trigger, menu selection, button selection, voice command;

---System automatically transfers to human. automatic transfer in specific scenarios, emotion perception.

5.2.2.2 Keyword Triggering Customers can directly input phrases such as "transfer to human," "human customer service," or "contact a human" in the chat interface with the intelligent customer service system to clearly express their preference. Once the intelligent customer service system recognizes the keywords for transferring to a human agent, it should automatically trigger the process of transferring the customer service to a human agent.

5.2.2.3 Menu Selection In the intelligent customer service chat interface, a "Transfer to human service" option or similar wording should be provided as a menu option. Customers can click this option to transfer the data to a human agent. You can then proceed to the process of being transferred to a live customer service representative.

5.2.2.4 Button Selection The customer service system should provide a button selection switching method. The button that triggers human customer service should maintain independent functionality within the service process to avoid [further issues]. It conflicts with other service buttons.

5.2.2.5 Voice Commands For customer service systems that support voice interaction, customer service representatives should use a standardized wake-up phrase, such as "Transfer to human agent" or "Transfer to human agent service." The system should recognize this wake-up phrase and... The call will proceed immediately.

5.2.2.6 Automatic call transfer in specific scenarios In the following specific scenarios, intelligent customer service should automatically transfer the call to a human customer service representative.

a) The customer contact center should set an interaction failure threshold. If the intelligent customer service system fails to reach a certain threshold for interaction with a customer regarding a particular product or service, the system will take action. If the threshold still fails to recognize the customer's expression, the call should be automatically transferred to a human customer service representative.

b) The customer explicitly states that they refuse intelligent customer service;

c) The conversation involved information security;

d) Emergency and risk scenarios involving personal and property safety;

e) Smart response timeout, for example, response time exceeds 3 minutes.

5.2.2.7 Emotion Perception After sensing and detecting negative emotions or complaint intentions in customers' interactions with AI-powered customer service, the customer service system, through communication with... Once the customer confirms, they should be automatically transferred to a live customer service representative.

5.2.3 Switching from human customer service to intelligent customer service

5.2.3.1 Switching Conditions When common customer service issues, strong customer preferences, heavy workloads, or emergencies arise, the customer should be clearly informed and their consent obtained. With the customer's consent, a switch from human customer service to intelligent customer service can be implemented.

5.2.3.2 Common Issues When customer inquiries are highly repetitive and these questions have clear answers in the intelligent customer service knowledge base, intelligent customer service... When customer service can provide quick and accurate answers, and with the customer's explicit notification and consent, human customer service may be provided. Switching to intelligent customer service.

5.2.3.3 Customer Preferences When customers show a clear preference for intelligent services, such as directly expressing "transfer to intelligent customer service" or similar statements, With clear notification and consent from customers, the switch from human customer service to intelligent customer service can be implemented.

6.1 Outbound Task Assignment

6.1.1 General Requirements The customer service system should assign work orders to either intelligent customer service or human customer service based on the complexity and type of the outbound service task. deal with.

6.1.2 General Scenarios When any of the following scenarios are met, the customer service system should prioritize assigning tasks to the intelligent customer service initiator.

---Closed-ended questions, such as those that can be answered through menu clicks, button selections, or simple voice responses;

---Standardization issues;

---Simple interactive services such as notifications and reminders.

6.1.3 Special Scenarios The customer service system should prioritize assigning tasks to human customer service representatives when one of the following scenarios is met. Outbound call services involve more complex tasks, such as handling personalized customer needs and ambiguous expressions;

---Outbound call services involve open-ended questions, such as customer follow-ups and market research that include open-ended questions and answers;

---Other outbound services that cannot be effectively completed by intelligent customer service.

6.2 Switching between outbound customer service calls

6.2.1 Switching from Intelligent Customer Service to Human Customer Service If the outbound call is initiated by the intelligent customer service system, the customer service system should provide the customer with an option to switch to human customer service, which can be done via keyword. Provide specific methods such as sending, menu clicking, button selection, and voice response. The specific switching should comply with the requirements of 5.2.2.

6.2.2 Switching from human customer service to intelligent customer service If the outbound call is initiated by a human customer service representative, the customer service system can provide an option to switch to intelligent customer service, which can be triggered by keywords. The specific methods of switching between menus, button selection, and voice response should be provided. The specific switching should comply with the requirements of 5.2.3.

7.1 End the task

7.1.1 Human Customer Service Customer service may be terminated under the following circumstances.

---Online customer service. After obtaining customer permission (e.g., the customer indicates "service can end") and the customer does not respond within 60 seconds of completing the service task. If a customer makes a new service request or remains inactive for more than 60 seconds, the online customer service representative can politely end the service.

---Human telephone customer service. After completing the service task and obtaining the customer's permission, ask the customer if they have any other questions, or initiate a live telephone customer service call 10 seconds later. If the customer still does not respond after two interactions, the human telephone customer service representative can politely end the service.

7.1.2 Intelligent Customer Service Intelligent customer service may terminate in the following circumstances.

---Intelligent Online Customer Service. After obtaining customer permission (e.g., the customer indicates "end service") and completing the service task within the threshold time, if the customer does not respond... If a new service request is submitted or a customer fails to interact for more than a certain period of time, the intelligent online customer service can automatically end the session. Organizations can adjust this based on... Set thresholds according to actual conditions;

---Intelligent Voice Customer Service. If the customer does not request new services or the customer exceeds the limit after 60 seconds of obtaining customer permission and completing the service task, or if the customer's request exceeds the limit... If there is no interaction after 60 seconds, the intelligent voice customer service will automatically end.

7.2 Service Closed Loop At the end of the service, the customer contact center should proactively invite customers to provide satisfaction feedback (if the customer declines, further disturbance should be avoided). (For each user), a complete service record is generated simultaneously, work orders are filled out, and pending tasks generated during the service process are circulated and closed through the work order system. Ring treatment.

8 Information Security

8.1 Information Backup Customer contact centers should establish customer information tiers based on core information such as customer identity information, historical interaction records, and incomplete work order data. The backup mechanism should combine real-time incremental backups with periodic full backups to prevent data loss due to a single storage failure. Specific information backups should... Please refer to GB/T 22239.

8.2 Classified Protection Customer contact centers should implement corresponding security measures in accordance with GB/T 22239, including but not limited to refined access control. Control and manage encrypted transmission and storage of backup data, backup operation logs, etc., to ensure the security and controllability of backup information throughout its entire lifecycle.

9 Collaborative Service Improvements

9.1 Improvement Mechanism Customer contact centers should consider both human and intelligent customer service as a complete service, and establish a system that integrates human and intelligent customer service. A service collaboration and improvement mechanism has been established, with designated personnel responsible for collaborative service improvement efforts. Regular surveys and the collection and processing of customer complaints and feedback are conducted. The handling of the lawsuit is based on GB/T 17242.

9.2 Improvement Methods Customer contact centers should collect relevant information to provide a reference for improving collaborative services between human and intelligent customer service. Specifically, they should collect... Information methods include, but are not limited to.

---Survey. Customer contact centers should regularly conduct surveys related to improvements in the collaboration between human and intelligent customer service. The relevant content is included as part of the customer service satisfaction survey;

---Complaints and Feedback. The customer contact center should have a complaint and feedback function for the collaborative service between human customer service and intelligent customer service. Channels, including but not limited to telephone, email, and online messages, should have established rules for handling public/customer complaints. and processing time;

---System Monitoring. The customer service system monitors interactions, customer service experiences, and dynamic feedback on customer requests;

---Third-party assessment. Customer contact centers can commission qualified third-party organizations to conduct assessments by simulating real customer interactions. An evaluation report was generated by analyzing interactive scenarios and breaking down key nodes in the collaboration between human and intelligent customer service.

9.3 Improvements Improvements to the collaborative services of human and intelligent customer service should include, but are not limited to, the following key aspects.

---Collaborative service efficiency. This includes the response speed and problem resolution rate of collaborative human customer service and intelligent customer service;

---Collaborative service quality. This includes customer satisfaction, service reliability, and personalized service quality in collaboration between human and intelligent customer service. Service capabilities, etc.;

---Smoothness of customer service switching. including the smoothness of switching between two directions;

---Customer Information Management. This includes the real-time synchronization and accuracy of customer information during the collaboration between human and intelligent customer service processes, and its storage. Storage security and information-based service continuity assurance.

......
This preview omits tables, figures, formulas and parts of the technical clauses. The complete document — 24 pages — is available in the English PDF.

How to Buy GB/T 47746-2026

1Add to cart. Click the "Buy GB/T 47746-2026" button on this page. You can add more standards before checkout.
2Checkout. Enter your email and billing details. Payment is processed securely by Stripe (cards, Apple Pay, Google Pay supported).
3Instant delivery (0–9 sec). Delivery is automatic: within seconds of payment you'll receive an email with a secure download link. The link stays valid for 72 hours.
4Invoice included. A tax invoice is attached to the confirmation email. Need a custom invoice? Contact us.