Agents Getting ‘Burned Out’ by Monotony of AI

Artificial intelligence was sold as a support tool for contact center agents. However, these days the agents working beside bots are finding themselves increasingly drained.

According to Omdia’s latest CX survey, many customer experience leaders in North America are now scrambling to figure out how to shield their employees from burnout.

A separate study by NiCE and CMSWire Insights reinforces the concern. It shows that 78% of supervisors worry that rising agent stress is damaging the customer experience.

The core problem, experts say, is that AI tools are behaving more like watchdogs than helping hands.

A large majority of contact centers have automated repetitive tasks. That leaves human agents to handle the toughest and most emotional calls.

According to Michael Moran, Vice President of IT at the nearshore outsourcing firm NQX, agents are burning out under the pressure.

Even small errors made during these intense calls are spotted instantly and escalated to managers.

Moran explains the change: “Previously, only a small sample of agent calls could be manually reviewed and evaluated. Today, automated QA solutions can analyze 100% of calls, identify problematic interactions, and uncover patterns that would be missed with limited sampling.”

Consider what happens in a routine high-stress situation. An angry customer calls. A bot takes the first shot at resolving the issue. When it fails, a human agent steps in. The customer then grows even more irritated, realizing they were talking to a bot just seconds earlier.

Agents are expected to use empathy to calm things down, but many have never been trained to express empathy in an environment where AI tools whisper instructions while monitoring every word. A single slip of the tongue can become a costly mistake.

Moran added that “real-time sentiment analysis can amplify the perception of a ‘Big Brother’ effect among agents.”

Therefore, Moran says, BPOs may now have to rethink how they measure performance. “More complex calls naturally mean higher average handle time and lower first call resolution. Their expectations and KPIs around these metrics should adjust accordingly.”

Training and Empathy Gaps

BPOs rushed to deploy AI to cut costs. But they skipped an essential step — training human agents to work alongside these systems, according to Joel Martins, CTO at Calabrio, which builds technology for contact centers.

Martins warns: “Too often, AI tools are introduced without the guidance and without a context.” He adds, “When agents don’t understand how AI makes decisions or how it benefits them, the tools feel intrusive rather than empowering.”

He cited a recent Harvard Business Review study showing that employees who received hands-on AI training reported 144% higher trust in workplace AI than those who did not. Meanwhile, Pew Research found that only 12.2% of workers have ever received AI-related training.

Martins argues that genuine AI success comes when agents see technology as an extension of their skills, not a threat to their autonomy.

Mark Speare, Chief Customer Success Officer at B2Broker, says CX automation is far more complicated, adding that AI-driven performance measurement has unintended consequences.

“AI can be used to analyze thousands of support calls, but this measurement has led to the development of unrealistic expectations for human workers, demanding precision and consistency in every interaction they have with customers.”

“Artificial intelligence operates on logic, but customer support isn’t a logical exchange between machines. Customers aren’t robots, and neither are the agents helping them.”