24 October 2026
AI is taking the easy calls, and leaving call-center staff with the hard ones
Companies are handing routine customer-service work to AI agents. Some staff say it frees them up; others say they now spend all day on the toughest, angriest cases.
Customer service is one of the first jobs where companies are putting AI agents to work at scale. That makes it an early test of a question every investor in AI should care about: when software takes over the simple parts of a job, does the human part get better, or just harder?
What happened
CNBC spoke to people on both sides. Kaylee Harris, who works in customer service at a credit card company, said AI now handles much of the note-taking and long card disclosures that used to fill her day, so she can give each caller more attention. A worker at a telecoms company in Wales told a different story: AI-written call notes were so full of errors that she writes her own, and at a previous job, AI booking agents handed anything they couldn't sort out to humans, leaving staff with all the complicated work and plenty of burnout.
Companies selling the technology say it can make a punishing job better. Nikola Mrkšić, chief executive of voice AI firm PolyAI, described contact-center work as "maximum abuse for minimum wage" and argued AI can share that load. UK telecom group Virgin Media O2 launched an AI voice agent in September to handle routine issues. ArvatoConnect, a UK customer-experience firm, aims to have 70% of interactions handled by technology and 30% by people, and says one tool it built for a UK government agency drafts supplier assessments with 97% accuracy.
Why it matters
There is a catch that researchers call work intensification: a job getting more demanding over time even if the hours stay the same. Amanda Jones, who studies it at King's College London, said the simple calls gave agents a breather between tough ones. Take those away and every call is the hard one, often with a customer already annoyed after dealing with a bot. Faster output can also mean tougher targets, and new staff may lose the easy cases they once learned on.
The data hint at the pressure. In a 2025/26 survey of UK contact-center managers by MaxContact, more than half said agent workloads had risen, up from around 40% in 2024, though the survey does not say why.
For investors, this is the gap between an AI demo and AI in the real world. Cost savings on paper can leak away through burnout, staff turnover and frustrated customers, which is why some firms argue human agents should be judged on whether they solved the problem, not how fast they hung up.
What to watch
- Whether companies relax call-time targets as humans take on harder cases.
- Customer satisfaction scores at firms that have rolled out AI agents.
- Staff turnover in call centers, an early sign of burnout.
Based on reporting by CNBC: AI in customer service: How jobs are changing. Facts, not advice.
More news
- Anthropic's IPO, in numbers: 825 million shares at $70 to $78, and a $1.4 trillion price tag
- Trump says the Nvidia scandal didn't endanger US security, and stands by Lutnick
- Nvidia's first earnings without Huang: what Kress will be asked on Tuesday
- Buffett gives away $5.5 billion of Berkshire stock in his Thanksgiving letter
- Banks are using AI to trim back offices. Proof that it adds revenue is harder to find
- IMF chief says the world economy leans too hard on AI as circular deals top $1 trillion