Customer-service workers describe AI relief—and extra work—in new CNBC interviews

For one customer-service worker, AI has made more room to focus on customers. For another, it produces notes she has to replace herself. In interviews published by CNBC on October 24, Kaylee Harris and Shannon describe sharply different experiences with workplace automation—and what happens to the work that stays with people.

Less paperwork, or another draft to fix

Harris works in customer service for a credit card company. Her day previously revolved around calls, note-taking and lengthy card disclosures. She told CNBC that AI now lets her concentrate more on each customer, rather than spending so much of the interaction on those accompanying tasks.

Shannon, a telecommunications customer-service worker in Cardiff, Wales, described the opposite result from automated note-taking. She said the notes at her current workplace often contain errors, “so I just write my own.” CNBC withheld her surname because she feared being identified by her employer.

Her experience also extends beyond notes. At a previous employer, a property-services company, AI agents booked appointments. When they could not offer a customer the requested date, they passed the case to human staff. Shannon said employees were “constantly fighting fires,” with complicated cases reaching them and causing burnout.

She also recalled senior managers at that employer telling staff to “ChatGPT it” when they did not know an answer. Her account describes both AI used by employees for help and automated agents handling customer requests before passing unresolved problems back to people.

The easy calls serve a purpose

Nikola Mrkšić, co-founder and CEO of customer-service voice AI developer PolyAI, sees automation as a way to ease a difficult job. He told CNBC that agents face frustrated customers and strict performance measures, sometimes without the information or authority needed to solve their problems.

But Amanda Jones, who researches organizational behavior and the future of work at King’s College London, warned that removing routine interactions can change the intensity of the remaining job. Customers may already be frustrated when they finally reach a person, she told CNBC.

Jones also identified a training problem: routine work gives employees experience that can prepare them for harder assignments. Automating it could remove those opportunities while leaving workers responsible for more demanding cases. She cautioned that productivity gains could also lead to tougher performance targets.

One thing we sometimes overlook is that relatively straightforward tasks can provide a bit of psychological respite between more demanding ones.

Amanda Jones, King’s College London researcher, speaking to CNBC

Harder cases need different measures

Workload pressure is already visible in a MaxContact survey cited by CNBC. More than half of surveyed U.K. contact-center decision-makers said agent workloads had increased in 2025/26, compared with around 40% in 2024. The survey does not establish what caused that rise, including whether AI contributed.

Debra Maxwell, CEO of customer-experience company ArvatoConnect, said average handling times may need to increase as people take on more complex cases. She described assessing whether agents understood customer needs, used sound judgment and reached the right resolution—not whether they followed a script or met an arbitrary time target.

Even Harris, who sees benefits despite occasional glitches, worries about the emphasis on speed. She told CNBC that choosing quantity over quality risks leaving customers fed up with AI. Shannon likewise sees potential benefits, but said it should not replace real customer service, questioning whether AI can make customers feel genuinely cared for.