Young People Driven Crazy by AI Customer Service: Finding a Human Agent Is Harder Than Climbing to Heaven
-
Trying to solve a problem with customer service, only to realize after a long conversation that the other party is an AI robot... Many netizens complain on social media that finding a human customer service agent nowadays is 'harder than climbing to heaven.'
Common complaints about AI customer service include 'talking past each other,' 'not understanding what's being said,' and 'wasting time.' 'Before, you could reach customer service with one click, but now you have to listen to an AI recite an 800-word essay first,' one netizen remarked. Others question: Is AI truly intelligent or just plain dumb? Despite widespread consumer criticism, why do so many companies still opt for AI customer service?
With the post-Double 11 after-sales wave approaching, many netizens complain about encountering numerous AI customer service agents on e-commerce platforms. 'They don’t understand what I’m saying, and their answers are irrelevant, which ruins my mood,' one user said.
'I asked when my order would ship, and it replied, ‘Dear, this question has stumped me.’ ‘When I was buying, they called me ‘sister’ and urged me to place the order, but during after-sales, they use cold AI to brush me off—knowing nothing and showing no emotion,’ another user lamented.
Some netizens claim that an e-commerce platform uses AI customer service disguised as human agents. "A few days ago, I couldn't claim an 1800-yuan coupon, so I contacted customer service. The page showed a human agent with an ID number, but it was actually AI. When I asked why the coupon failed, it replied, 'Our discount is really huge this time,' completely missing the point!"
Beyond e-commerce platforms, AI customer service is also widely used on websites, ride-hailing apps, and even in the courier industry.
One netizen shared their experience with Apple's customer service: "After calling Apple's support line, I had to deal with AI for three minutes and was forced to listen to a song before finally reaching a human agent. It was such a waste of time."
Others criticized J&T Express: "To reach their official human service, I had to battle their dumb AI customer service for ages."
▲ Source: Social media screenshot
Some netizens reported similar issues with ride-hailing apps. "Once I booked a ride for just a 1-kilometer distance, waited over 10 minutes for the driver who never arrived, and when I contacted customer service, all dozen calls were answered by AI."
Others complained about confusing social media platform support. "I asked customer service if there was a group chat red envelope reminder feature, and the AI replied with 'How to set up Do Not Disturb mode.'"
Additionally, many telemarketing calls have been replaced by AI robots, particularly in finance, real estate, logistics, communications, and tourism industries. One user noted: "Now even property sales calls use AI that mimics human voices perfectly. But when you ask questions, it's like talking to a brick wall - they just keep reciting their script regardless of what you say."
As early as 2019, CCTV's 3·15 program reported the rampant use of AI in making spam calls. While a human can make only 200-300 calls a day, a robot can make up to 5,000. Some companies even made over 4 billion spam calls in a single year.
Behind the widespread adoption of AI customer service across industries is the lack of human customer support. When users encounter problems and seek help, whether through online platforms or phone calls, AI customer service often stands in the way. During communication, these systems either provide a list of pre-set answers or slowly read out scripted responses. If the user's query falls outside these predefined options, the conversation becomes disjointed, leaving many consumers frustrated by the difficulty in communication.
A recent investigation by China News Service found that some popular e-commerce stores use AI customer service. For common questions like "delivery time," "shipping," "size," or "color discrepancy," the AI typically responds by triggering keywords. However, when users ask for details or change keywords, the system reacts unpredictably—sometimes transferring to a human agent, other times providing irrelevant answers.
On the 15th, Zhongxin Jingwei sent a link of a fleece jacket from a sports brand's online flagship store to customer service, asking questions such as "Can this be worn in Beijing during winter?", "Does the black color fade?", "Will the fabric pill?", and "What material is this made of?". The AI customer service responded to all these queries with relatively accurate answers.
However, when asked "Why does it look so thin?", the response switched from AI to a human agent. Upon inquiring about the working hours of human customer service, the representative stated, "Human agents are available from 9:00 to 24:00." Yet, when questioned why AI responses were still used during these hours, the reply was "Not sure."
Additionally, Zhongxin Jingwei consulted a down jacket specialty store, asking whether the "letters on a particular down jacket were in English" and "what they meant." The AI customer service responded with, "This product features triple-lock warmth retention, is waterproof, stain-resistant, oil-proof, windproof, and an excellent choice for extreme cold weather," failing to address the actual query.
Moreover, reaching human customer service on some social media platforms has become increasingly difficult.
Ms. Mu from Beijing recently told China News Service that when she tried to consult customer service on a social platform about an issue, the AI chatbot responded with 14 irrelevant messages covering topics like membership, account anomalies, and lucky draws—none of which addressed her problem. She couldn't reach a human agent online at all.
When she later called the customer service hotline, an AI voice assistant repeated the same 14 automated responses in a loop. Attempting to transfer to a human agent, she was told "All agents are currently busy. You can schedule an online callback with a human agent." Ms. Mu discovered the earliest available callback slot was the next afternoon, forcing her to abandon the inquiry entirely. This back-and-forth wasted an hour without ever connecting her to human support.
On consumer complaint platforms, there are nearly 100,000 grievances related to AI customer service and chatbots, with most reflecting issues such as "human agents ignoring customers," "robots pretending to be human," and "inability to communicate effectively—just fooling people."
One netizen reported calling customer service lines of 45 companies, including telecom providers, e-commerce platforms, and logistics firms. The statistics showed that the average wait time to reach a human agent was 94 seconds, with the longest wait lasting up to 380 seconds.
The 2023 China Intelligent Customer Service Market Report indicates that China's intelligent customer service industry is dominated by AI customer service software, accounting for about 80% of the market share. In 2022, the market size reached 6.68 billion yuan, and it is projected to grow to 18.13 billion yuan by 2027.
With technological advancements, the market size of AI applications in the customer service sector continues to grow. How does the cost of AI customer service compare to that of human agents? What are its strengths and weaknesses?
Chen Yinjiang, Deputy Secretary-General of the Consumer Rights Protection Law Research Association of the China Law Society, stated that for businesses, the promotion of intelligent customer service is often more about reducing costs. Internet transactions involve massive volumes and instantaneous processing, sometimes requiring substantial customer service resources. Appropriately implementing intelligent customer service can significantly enhance service efficiency. However, many intelligent customer service systems are cumbersome and often provide irrelevant answers, making it difficult to meet consumers' actual needs. Therefore, the key is to retain human customer service simultaneously, allowing consumers to choose their preferred service method.
Hu Gang, Deputy Secretary-General of the Legal Work Committee of the Internet Society of China, also believes that intelligent customer service can operate 24/7, drastically reducing labor costs, which is beneficial for businesses. However, for consumers, this may result in substandard after-sales service quality.
Additionally, the 2023 China Intelligent Customer Service Market Report mentions that current AI customer service solutions have limited problem-solving capabilities and receive slightly lower recognition than human operators. There remains a significant gap from achieving true intelligence, with repetitive responses and circular operations being the main pain points.
Regarding consumer complaints about some platforms' human customer service being "unreachable," does this business practice potentially violate consumer rights?
Chen Yinjiang believes that enterprises cannot only consider their own costs and efficiency while disregarding consumers' time. When consumers contact customer service, they usually encounter problems that need resolution. However, some companies focus all their energy on market competition without maintaining proper customer service personnel, using AI customer service as a "shield" that proves completely ineffective and serves no real purpose. "Although it's difficult to quantify the exact losses to consumers, it certainly creates poor consumer experiences. For businesses, reducing operational costs sometimes means increasing consumers' communication costs."
Can intelligent customer service completely replace human operators in the future?
Hu Gang believes that AI customer service will find it difficult to completely replace human agents in the future. "Customer service not only provides users with knowledge, information, and access to related products and services, but also crucially resolves disputes, which requires differentiated approaches to various issues—a relatively complex task. A key aspect of customer service is the need to comfort and soothe consumers' or clients' emotions, something AI struggles to achieve."
Chen Yinjiang also believes that AI customer service is unlikely to replace human agents. He suggests that enterprises adopt a 'human + AI' customer service system, where routine inquiries can be handled through automated responses, while urgent issues such as consumer complaints—especially those involving personal or property safety—should be prioritized for resolution through human customer service channels.