Wang Jie worked in customer service at an e-commerce company for eight years. She worked her way up from frontline support to supervisor, managing a team of more than twenty people. She knew every product, every common complaint, every escalation path by heart. She thought she'd retire there.
Then last month, her entire department was "cleaned out in one pot," as she put it. The company deployed an AI customer service system. Of the more than forty people in the customer service department, only three were kept on — and only to handle complex complaint cases that the AI couldn't resolve.
"The machine doesn't pay social insurance, doesn't get emotional, works 24/7 — the boss calculated it all," she said, laughing bitterly. "I did this for eight years, and I thought I was good at my job. But being good at repetitive work isn't the same as being valuable."
The numbers back her up. Industry data shows that AI customer service can already resolve over 85% of common issues, at one-tenth the cost of human agents. After one major Chinese bank rolled out smart customer service, they saw 2 million fewer phone calls per month. Another bank's AI bot has the service capacity equivalent to 9,000 human agents.
Tracking orders, processing returns, changing addresses — for these standardized problems, machines are faster and never get impatient. What used to take teams of dozens or hundreds of people now requires just a handful to handle the edge cases.
Wang Jie isn't sitting around waiting. She enrolled in a community operations training program, planning to pivot toward user operations. "I've been in customer service for eight years," she said. "I know better than anyone what customers are thinking. That's experience a machine doesn't have."
She's not the only one. Across China, an estimated 600,000+ positions in customer service, telemarketing, and basic administrative roles are being reduced this year alone. The first wave of AI elimination isn't hitting factory workers or delivery drivers — it's hitting people who sit in offices doing what everyone thought was "stable" white-collar work.