Platform Customer Service and Communication Compliance Management Policy

Effective Date: -06-01 | Scope: Omnichannel Customer Service System and Third-Party Service Providers

This "Platform Customer Service and Communication Compliance Management Policy" (hereinafter referred to as the "Policy") aims to regulate the customer service practices of (hereinafter referred to as the "Platform"). While ensuring users receive an efficient and professional service experience, this Policy strictly prevents customer service channels from being misused as tools for telecom fraud traffic diversion, illegal information dissemination, embezzlement concealment, money laundering communication, or malicious theft of trade secrets. Clauses involving service restrictions, privacy protection, recording retention, and legal liability are highlighted in bold or underlined; please read them carefully.

I. Service Standards and Response Time Commitments

1. Multi-Channel Service Matrix
   • Provision of 7×24h AI + Live Chat support, weekday 9:00–21:00 telephone hotline, ticketing system, and email support;
   • VIP users and enterprise clients are entitled to dedicated account managers and priority access channels.
2. Response Time Benchmarks
   • First response time for live chat ≤30 seconds; average handling time ≤8 minutes;
   • Hotline connection rate ≥95%; wait time ≤60 seconds;
   • Standard inquiries via tickets/email answered within 24 hours; complex issues resolved with a solution plan within 48 hours and continuously followed up.
3. Service Quality Red Lines
   Customer service personnel are strictly prohibited from using insulting, discriminatory, or threatening language; strictly prohibited from disclosing user privacy; strictly prohibited from privately adding users' personal social media accounts; strictly prohibited from guiding users to transact outside the platform; and strictly prohibited from evading responsibility for reasonable user requests. Violators of these red lines will be immediately terminated upon verification and held legally accountable.

II. Anti-Fraud and Prohibitions Against Abuse of Customer Service Channels

1. Zero Tolerance Stance
   The Platform strictly prohibits any user or internal personnel from engaging in the following illegal or criminal activities via customer service channels:
   • Impersonating customer service to conduct telecom fraud, send phishing links, or extract verification codes;
   • Transmitting coded language or steganographic messages via customer service dialogues for illegal transactions such as gambling, drugs, or pornography;
   • Fabricating complaints or emergencies to induce customer service to bypass risk control rules for account unfreezing, password modification, or fund transfers;
   • Coercing customer service to provide merchant backend data, other user information, or core platform algorithms under the guise of "rights protection" or "complaints";
   • Initiating bulk invalid inquiries to occupy resources and disrupt normal service order (i.e., "customer service bombing").
2. Abnormal Communication Risk Control Monitoring Model
   The Platform deploys real-time semantic analysis and behavior recognition systems to automatically trigger alerts for the following scenarios:

   Alert Dimension	Specific Trigger Scenario	Risk Level	Automated Action
   Sensitive Keyword Hit	Dialogue contains high-risk words like "transfer," "safe account," "verification code," "click farming"	Extremely High	Real-time pop-up warning to user + Session recording flagged + Transfer to anti-fraud specialist
   Abnormal Emotion/Speech Rate	Continuous abuse/threats of self-harm by user, or sudden speech rate changes/prolonged silence by agent	High	Supervisor real-time monitoring intervention + Psychological crisis intervention protocol activation
   High-Frequency Repetitive Inquiry	Same user initiates ≥20 identical inquiries within 24h	Medium	Throttled to AI bot + Secondary identity verification required
   Cross-Channel Correlation Anomaly	Phone, chat, and tickets simultaneously reflect the same non-genuine request	High	Merged session analysis + Suspension of non-essential permission operations
   External Traffic Diversion Attempt	Appearance of WeChat ID, QQ number, external links, QR codes, etc., in dialogue	Extremely High	Automatic content blocking + Account ban + Report to Cyber Police
   Internal Compliance Violation	Agent frequently queries user data outside job scope or exports records in bulk	Extremely High	Immediate operation blockage + Work ID freeze + Internal audit investigation launch

   ⚠️ Important Notice: The above model is dynamically iterated. Users may not claim service rights on the grounds of "not triggering published rules." The Platform reserves final discretion to terminate services or implement control measures based on reasonable suspicion.

3. User Identity Verification Obligation
   For sensitive operations involving account security, fund transactions, or privacy information queries, customer service has the right to require multi-factor authentication (e.g., SMS verification code + facial recognition + historical order verification). The Platform has the right to refuse service without liability for delays if users refuse to cooperate or fail verification.

III. Prevention of Occupational Crimes and Interest Transport via Customer Service

1. Prohibition of Collusion and Rent-Seeking
   The Platform strictly prohibits customer service personnel, management, or third-party service providers from abusing their positions to:
   • Accept bribes from users/merchants to illegally elevate service priority, delete negative reviews, unblock accounts, or leak competitor data;
   • Fabricate user complaint records to fraudulently obtain platform compensation funds or performance bonuses;
   • Use customer service permissions as leverage to secure unfair competitive advantages for specific related parties;
   • Participate in insider trading, extortion, or sell data to black/grey market industries using sensitive information at hand.
2. Dual-Person Review Mechanism for High-Risk Operations
   • All operations involving fund compensation >500 RMB, account unblocking, password resets, or data exports must be secondarily authorized by supervisor-level or higher personnel with approval logs retained;
   • Service records for enterprise client dedicated agents shall be spot-checked and audited monthly by the Compliance Department, focusing on abnormal preferential treatment.
3. Clue Referral and Judicial Cooperation
   Upon discovery or reasonable suspicion of crimes such as bribery, infringement of citizens' personal information, or destruction of computer information systems in customer service operations, the Platform will unconditionally transfer all call recordings, chat logs, operation logs, and fund flow records to public security organs and supervisory commissions in accordance with the law, without prior notice to the involved parties. The Platform assumes no liability for breach of contract due to service interruption or delay caused by cooperation with investigations.

IV. Emerging Customer Service Risk Alerts and Emergency Response

In response to high-frequency risk scenarios in , the Platform has formulated special contingency plans:

• "AI Voice Cloning" Impersonation Fraud: Criminals synthesize customer service voices to call users and induce transfers.
  Countermeasure: Official outbound calls uniformly use certified numbers + dynamic voiceprint watermarks; users may request the agent to state the session-specific verification code at any time; suspected cloned calls are automatically flagged with anti-fraud alerts pushed.
• "Deepfake Video" Bypassing Facial Verification: Using AI face-swapping technology to deceive agents into completing identity verification.
  Countermeasure: Upgraded liveness detection algorithms with added random action commands + ambient light reflection analysis; high-risk operations mandatorily converted to offline outlets or notarized processing.
• "Public Opinion Hostage-Taking" Malicious Complaints: Organizing astroturfing campaigns to post false negative service reviews on social platforms to coerce agents into meeting unreasonable demands.
  Countermeasure: Establish a joint public opinion-complaint assessment mechanism to distinguish genuine feedback from malicious attacks; verified extortion cases will have evidence preserved and be referred to the Special Task Force on Combating Organized Crime.
• "Social Engineering Database" Precision Script Information Extraction: Using leaked data to fabricate realistic scenarios to gain agent trust.
  Countermeasure: Real-time knowledge graph validation embedded in the CS system; automatic alerts triggered when dialogue contradicts user profiles; additional dynamic security questions required for sensitive information queries.

V. Data Retention, Privacy Protection, and Violation Handling

1. Full-Volume Service Record Retention
   All call recordings, online chat logs, ticket workflow logs, and operation audit trails shall be retained for no less than 2 years from the date of service completion; for cases involving disputes, complaints, or judicial investigations, the retention period extends to 5 years after case closure. Storage employs national cryptographic algorithm encryption + blockchain hash anchoring to ensure immutability and traceability.
2. Privacy Minimization Principle
   Agents may only view necessary information directly related to the service; sensitive fields (ID numbers, bank card numbers, phone numbers) are masked by default. Screenshots, screen recordings, transcription, or verbal relay of user privacy are strictly prohibited. Violators will be referred to judicial authorities for the crime of infringing citizens' personal information.
3. Violation Handling Measures
   • User Side: Warning, throttling, service suspension, permanent ban, inclusion in industry shared blacklist, referral to judicial authorities;
   • Employee/Service Provider Side: Warning, demotion, termination, damage recovery, referral to judicial authorities, cooperation termination, and deposit deduction.
4. Liability Exemption
   The Platform assumes no liability for compensation for service restrictions, delays, or terminations caused by user illegal acts, identity verification failures, force majeure, or risk control measures taken by the Platform in accordance with regulations.

VI. User Remedies and Appeal Mechanisms

1. Scope of Appeals: Only appeals regarding service issues caused by system misjudgment, poor agent attitude, or operational errors are accepted; no appeal channel is provided for risk control measures taken based on reasonable suspicion.
2. Appeal Material Requirements: Applicants must submit proof of identity, service session ID/timestamp, problem description, and supporting materials (e.g., screenshots, audio clips).
3. Appeal Processing Timeline: For complete submissions, the Platform will issue a review decision within 3 business days; if supplementary materials are required, the timeline restarts from the date of completion. Original control measures remain effective during the appeal period.
4. Finality Statement: The Platform's review decision is final. Users who still object to the result should seek resolution through judicial channels and shall not interfere with the Platform's normal operations through repeated appeals, petitions, or public opinion pressure.

Appendix: Implementation Execution and Technical Support Checklist