The Shanghai Consumer Protection Commission recently revealed that it received over 320000 consumer complaints last year. Among them, the "non intelligence" of intelligent customer service has become one of the hot issues in the field of consumer rights protection. Currently, more and more platforms and apps have launched intelligent customer service, but many of them have low processing efficiency and fail to answer questions, which has become an obstacle to after-sales rights protection. The reason for this is, on the one hand, that businesses are continuously reducing labor costs and even completely canceling manual customer service, pushing all after-sales services to artificial intelligence; On the other hand, due to technological limitations, intelligent customer service with limited intelligence is unable to cope with personalized consulting needs. These issues need to be taken seriously and improved. Firstly, enterprises should strengthen their sense of responsibility and continuously optimize the level of intelligent customer service. Intelligent applications should not deviate from the original intention of the service, let alone layer by layer transfer and avoid problems. At the same time, considering the practical needs of diversity and adaptability to aging, manual customer service cannot be "absent". Overall, it is necessary to provide customer service types that consumers can choose independently, ensure that the after-sales force matches the market size of the enterprise, and protect the various rights and interests of consumers. Secondly, intelligent customer service needs to enhance its ability to solve diverse problems. To achieve this, mature artificial intelligence technology is needed as support. There are currently many types of intelligent customer service models on the market, but the cost of long-term training models is relatively high. The combination of industry, university, research and application can be used to reduce the application cost of new technologies, better optimize the language model of intelligent customer service, effectively improve the ability to identify and respond to changes, and improve the service level. Finally, industry organizations should establish unified institutional norms and service standards as soon as possible. For example, establish a service evaluation scoring mechanism and introduce labor standards, technological ethics, and other aspects into the evaluation standards. We will severely punish individual merchants who use intelligent customer service to avoid users' after-sales needs due to communication difficulties. In summary, all parties involved should make every effort to return intelligent customer service to its true service nature, answer questions more accurately and reliably, and protect consumer rights. It is also important to be aware that, limited by the current level of technology, intelligent customer service cannot completely replace manual customer service. Currently, a human-machine collaboration approach that relies mainly on manual labor and supplemented by technology may be a feasible direction to address the pain points of intelligent customer service and continuously improve customer service work. (New News Agency)