AI in customer service: burnout on one side, no measurement on the other

In the United States, CNBC finds customer service workers split. Some say AI frees them from admin so they can handle complex cases; others report error-prone generated notes and burnout, because automation leaves them only the hardest calls, and the easy ones, which used to be a breather, disappear.

Other markets are at an earlier and calmer point. A survey of about 100 Swiss firms shows that 36 percent already use AI in customer service and 41 percent are implementing it; 56 percent see it as a complement to the team and only 8 percent expect large-scale replacement (inside-it). But 46 percent do not measure their AI agents at all, and only about 20 percent can quantify the return.

The two stories describe the same risk from different ends. When the machine takes the easy work and nobody redesigns the job, people burn out. When nobody measures, nobody notices. Supporting people with AI only works if the human role is designed on purpose, with targets, breaks and training adapted to the new mix of calls.

Customer service employs millions of people, and it is one of the first fields to change. The agents who learn to supervise and correct AI keep their jobs; those who are not offered that training are the first to go. Your donation gives them that training.

Practical step: before the next rollout, define three numbers (resolution rate, handling time, load on staff), track them monthly and adjust the human role.