Compliance one-pager · single A4 sheet, print and tick Download PDF · full evidence & online self-test: compliance-check.html

GB/T 47746-2026 Compliance One-Pager Self-Check List

GB/T 47746-2026 “Customer contact service — Requirements for the collaboration between human and intelligent customer service” (issued 2026-05-25, effective 2026-09-01, a recommended national standard, administered by SAC/TC 264) · 61 items in total (shall 48 / should 4 / may 9, covering 31 clauses)

How to use: check the 5 veto items below first — if any one of them fails, the rest do not make you compliant. Then tick your way through the six groups; an unticked box is a gap.

⚠ Veto items (5.2.2.6, all “shall”) — failing any one means non-compliant

  • 5.2.2.6 a)Interaction reaches the threshold and the customer’s expression still cannot be recognised
  • 5.2.2.6 b)The customer explicitly declines intelligent customer service
  • 5.2.2.6 c)The conversation involves information security
  • 5.2.2.6 d)An emergency or risk scenario affecting personal or property safety
  • 5.2.2.6 e)Intelligent response times out (the standard’s example: over 3 min)

1. Organisation & governance (4.1–4.4)

  • A1shallStaff human customer service matched to your industry and user demand — you cannot leave only an AI with nobody behind it4.1
  • A2shallEstablish an organisational policy for collaboration between human and intelligent customer service4.1
  • A3shallDefine contingency mechanisms for disasters, system failures and similar events4.1
  • A4shallDefine the organisational structure for human–AI collaboration4.1
  • A5shallBuild an internal system covering customer information management and ticket handling4.1
  • A6shallProvide the customer service entry point in a clear, prominent and unambiguous way4.1
  • A7shallEstablish a knowledge base4.1
  • A8shallRun regular optimisation training for the intelligent service and train customer service staff4.1
  • A9shallClear division of duties — the intelligent service takes standardised questions that have a definite answer in the knowledge base4.2
  • A10shallHuman agents take questions requiring multi-step analysis or several information sources, plus anything the intelligent service could not solve4.2
  • A11shallSeamless two-way switching — the system must pass customer identity, interaction history and other key data across4.3
  • A12shallEstablish information and data security mechanisms covering collection, processing and use, plus model and personnel management4.4

2. Service start & routing (5.1)

  • B1shallThe service system must provide the customer service entry point clearly, prominently and unambiguously5.1.1
  • B2shallAt the start, allocate tickets according to problem-solving capability, overall call volume and capacity, and the customer group's acceptance of intelligent service5.1.1
  • B3shallTell the customer the queue position or expected waiting time by text or voice5.1.1
  • B4shouldSimple, repetitive or standardised questions, questions with a definite knowledge-base answer, and high-volume periods where human agents are overloaded — the intelligent service is preferred5.1.2
  • B5shallThe customer explicitly asks for a human agent — the human agent is preferred5.1.3
  • B6shallComplex questions (individual needs, vague phrasing, cross-department coordination) — the human agent is preferred5.1.3
  • B7shallRare questions (the standard's example: appearing no more than 3 times in the past year) — the human agent is preferred5.1.3
  • B8shallThe knowledge base has no definite answer — the human agent is preferred5.1.3
  • B9shallScenarios involving the customer's personal or property safety — the human agent is preferred5.1.3
  • B10shallThe customer has come back 3 or more times within 24 hours and is clearly dissatisfied with the intelligent service — the human agent is preferred5.1.3
  • B11shallThe customer is in a specific age group (elderly, minors) or has low acceptance of intelligent service — the human agent is preferred5.1.3
  • B12shallThe customer is extremely emotional or threatens extreme behaviour, beyond what the intelligent service can handle — the human agent is preferred5.1.3

3. Human–AI switchover (5.2)

  • B13shallWhen switching, key data must be transferred completely and accurately: customer identity, interaction history, problem description and ticket status5.2.1
  • B14shallDuring the wait before the switch, tell the customer the expected waiting time by voice or text5.2.1
  • B15shouldOnce the switch is complete, the customer should not have to repeat information already given or restate the problem5.2.1
  • B16shallKeyword trigger — recognising “human agent”, “talk to a person”, “contact a human” and similar phrases triggers the transfer automatically5.2.2.2
  • B17shallMenu option — the conversation interface should offer “transfer to human service” or similar as a menu item5.2.2.3
  • B18shallKey press — the key that triggers a human agent must stay functionally independent and not conflict with business keys5.2.2.4
  • B19shallVoice command — voice-enabled systems should announce a consistent wake word (e.g. “human agent”) and transfer immediately once recognised5.2.2.5
  • B20shallAutomatic transfer — the system needs scenario-based automatic transfer and sentiment awareness5.2.2.1
  • B22shallSentiment awareness — after detecting negative emotion or intent to complain, and after confirming with the customer, transfer automatically to a human5.2.2.7
  • B23mayHuman → intelligent switching — the customer must be clearly informed and must consent (four situations: FAQ, customer preference, heavy workload, emergency)5.2.3.1
  • B24mayHuman-to-AI handover (repetitive questions): when the question is highly repetitive and the knowledge base has a clear answer that can be given fast and accurately, handover is allowed — but you must inform the customer and get consent first5.2.3.2
  • B25mayHuman-to-AI handover (customer preference): when the customer explicitly prefers AI service (e.g. asks to be transferred to the AI assistant), handover is allowed — with prior notice and consent5.2.3.3

4. Outbound scenarios (6.1–6.2)

  • C1shallAllocate outbound tasks to intelligent or human service by complexity and type6.1.1
  • C2shouldClosed-ended questions, standardised questions and notification-type calls — intelligent outbound calling is preferred6.1.2
  • C3shallComplex tasks (individual needs, vague phrasing) and open-ended Q&A (customer follow-up, market research) — human outbound calling is preferred6.1.3
  • C4shouldIntelligent outbound calls should offer a route to a human (keyword / menu / key press / voice)6.2.1
  • C5mayHuman outbound calls may offer a route to the intelligent service6.2.2

5. Service closure & follow-up (7.1–7.2)

  • D1mayEnding conditions for live online service: customer consent / no new request 60 s after completion / no interaction for over 60 s7.1.1
  • D2mayEnding conditions for live phone service: completion with consent, then no response 10 s after asking “anything else?”, or after 2 follow-up attempts7.1.1
  • D3mayIntelligent voice service: may end automatically after consent / no new request 60 s after completion / no interaction for over 60 s7.1.2
  • D4mayIntelligent online service: the organisation sets its own threshold duration7.1.2
  • D5shallAt the end of service, proactively invite a satisfaction rating (and avoid further disturbance if it is declined)7.2
  • D6shallGenerate a complete service record and ticket at the same time; to-dos must flow through the ticketing system and close the loop7.2

6. Data & security (8.1–8.2)

  • E1shallEstablish tiered backup for customer identity, interaction history and open-ticket data; combine real-time incremental backups with periodic full backups8.1
  • E2shallApply the security protection measures for the corresponding level under GB/T 22239: fine-grained access control, encrypted transfer and storage of backups, and backup operation logs8.2

7. Continual improvement (9.1–9.3)

  • F1shallTreat human and intelligent service as one service; establish a collaboration improvement mechanism with a named owner, and regularly collect and review complaints and feedback9.1
  • F2shallHandle complaints in accordance with GB/T 172429.1
  • F3shallCarry out regular collaboration improvement surveys (these may be merged into satisfaction surveys)9.2
  • F4shallOpen complaint and feedback channels for the collaboration service (telephone, email, web form) with defined handling rules and time limits9.2
  • F5shallSystem monitoring — monitor interaction actions, agent experience and dynamic feedback on demands9.2
  • F6mayThird-party assessment — you may commission a body with professional customer-service qualifications to simulate real interactions, break down the key collaboration points and issue an assessment report9.2
  • F7shallImprovement must cover four areas: collaboration efficiency, service quality, switching smoothness, and customer information management9.3
Scope note: this list is compiled by SavantCat from the text of the standard. It is not published by the standards body and does not constitute certification, an audit conclusion or legal advice; the official text governs interpretation. Clause quotations are the shortest excerpts needed for locating the requirement.
Evidence & online self-test: https://savantcat.cn/en/compliance-check.html (tick all 61, get a gap report)   Open data: checklist-v2.0.json / .csv  For agents: MCP (self_check_list returns all 61)  Compiled by: SavantCat  Version: v2.0 (verified 2026-09-19)