As artificial intelligence is rapidly being adopted across customer service operations, the role of agents is shifting from simple responses to resolving complex problems.

CNBC reported on Oct. 24 that companies are moving to a model in which AI handles repetitive tasks while people focus on consultations requiring judgment and empathy.

Reaction in the field is mixed. Callie Harris, who works in a customer service department, said she used to spend much of her time handling phone calls, taking notes and explaining terms and conditions. Now she can focus more on each customer. With AI taking over administrative work, she said she can spend more time on the quality of consultations.

Shannon, a customer service worker at a telecommunications company in Cardiff, Wales, said notes generated by AI contain so many errors that she has to rewrite them each time. At a real estate services company where she previously worked, the response was along the lines of "If you don't know, ask ChatGPT." She said an AI customer service agent would schedule appointments but, if it could not match the date a customer wanted, it would ultimately hand the case to a person. As complicated cases piled onto people, she said, burnout worsened.

Companies see AI automation as a way to improve the efficiency of customer service operations. Nikola Mrksic, co-founder and CEO of London-based voice AI developer PolyAI, pointed out that customer centre agents often directly bear customer complaints amid low pay and performance pressure. He said AI could help create a better environment for both customers and employees.

Adoption is continuing in practice. British telecoms and media company Virgin Media O2 introduced an AI voice agent in September. The system is designed so AI handles repetitive inquiries and human agents focus on complex or sensitive issues.

The problem is that the more AI takes over simple tasks, the more difficult work may concentrate on human agents. A report by MaxContact, titled the 2025 to 2026 UK contact centre key performance indicator benchmarking insights report, found more than half of surveyed decision-makers said agents' workloads had increased. The figure was higher than about 40 percent in 2024, but the report did not identify the cause of the increase.

Amanda Jones, who studies organisational behaviour and the future of work at King's College London, warned that AI could increase the intensity of work in call centres. She said customers may already be more dissatisfied by the time they are connected to a human agent. She also said when simple tasks disappear, there are fewer chances to ease psychological pressure between high-intensity duties. She also pointed to the possibility that productivity gains could lead to higher performance targets, or that automating repetitive tasks could reduce opportunities for employees to learn their jobs.

Some also argue that standards for evaluating agents' performance need to change. They say performance should not be judged solely on whether agents followed a set script or met arbitrarily set time targets.

Frontline workers acknowledged the usefulness of AI but voiced concern that the customer experience could worsen. Harris said system errors can increase work at times, but overall it has advantages. She added that expanding automation is ultimately moving in a direction that sacrifices customer satisfaction.

As automation expands, responding to system errors, easing the burden on agents from difficult complaints and improving performance evaluation standards have also emerged as challenges. Whether AI improves consultation quality or increases agents' burden and customer complaints will depend on how companies apply and operate it in the field.