As companies automate simpler customer inquiries, customer service agents are handling more complex, emotionally charged work.

The result is a different kind of service job — one that requires more judgment, more emotional control and more real-time decision making.

“The routine issues are automated or removed or deflected,” said Jonathan Schmidt, senior principal analyst for customer service and support leaders with Gartner. “Agent roles have shifted from more or less execution to more judgment-oriented.”

While AI can reduce some of that burden by improving access to knowledge, summarizing calls and automating manual tasks, agents are increasingly handling edge cases that require judgment or human touch. It also raises the risk of higher employee burnout and declining customer experience quality, as agents are now, in some cases, moving from one complicated issue to another without a mental reset. 

“Today, in some extreme cases with clients, you’re getting edge case after edge case after edge case,” said Tim McDougall, managing director and leader of Deloitte’s contact center practice.

It may come as no surprise that 60% of employees don’t want to take on more complex tasks, according to research from Gartner. 

More complex interactions can increase handle time, which can, in turn, extend wait times and affect staffing requirements. In an “always on” world, customers have limited patience for delays, especially when they have already tried self-service before reaching a person, McDougall said.

The result: Representatives are all too often faced with frustrated customers as soon as they pick up the line.

Too much information, too little context

Contact centers face another problem: Agents are not always missing information. Sometimes, they have too much information without enough context.

Agents may have access to customer data, scripts, recommendations and prompts, but still lack clear signals about what matters for the specific interaction, Schmidt said. Fragmented technology environments can force them to move across multiple systems to assemble a full picture.

But many organizations are still dealing with fragmented, legacy systems that don’t integrate well, making it harder for agents to access and act on information. As companies built hundreds of disconnected applications over time, agents effectively became responsible for stitching everything together.

“The customer service worker got the shaft and had to sort through all this data and all these systems,” said Nate Brown, co-founder and executive director of CX Accelerator.

Many companies are now repeating that pattern with AI, layering new tools on top of siloed systems — a move that can increase strain on agents and hurt customer experience.

The burden grows when information is inaccessible, incomplete or poorly organized.

“Information is the lifeblood of the work,” Brown said. “Where it becomes overwhelming is when the information isn’t there, when there are gaps in the information, or when it’s not properly organized.”

That distinction matters for AI implementation.

“AI reduces cognitive load only when it pairs the relevant context with the actual guidance,” Schmidt said. Otherwise, poor implementations can “filter complexity towards the agents instead of eliminating it.”

AI can be powerful when customer data and knowledge bases are consolidated, Experience Investigators CEO Jeannie Walters said. But if companies do not structure data correctly, agents may still have to search across multiple places while also learning and supervising new tools.

Tools that surface relevant information without requiring manual search can reduce cognitive load, McDougall said.

But many implementations focus on guidance without enough context, requiring agents to evaluate, override or justify system recommendations, Schmidt said.

How agent overload can diminish service quality

Juggling multiple channels and real-time prompts can erode agent confidence during live interactions.

Customers feel the effects, too. Overloaded agents may rely more on scripts, making interactions feel less empathetic, Walters said. 

Customers pick up on how agents feel, Brown said. “If the service agent feels helpless because the tools are broken, the data is not there or there’s just too much going on, it’s too hard, and the customer is going to feel helpless.”

Empathy is finite, and agents are being asked to show empathy even as their own decision fatigue grows, Walters said. If AI handles easier tasks, agents may lose the lighter interactions that once gave them a break.

High cognitive load can lead to emotional fatigue and reduced confidence, even among experienced agents, Schmidt said. Gartner research has found that technology environments with lots of guidance but limited context are associated with higher turnover intent, he said.

The issue is also financial. Replacing a skilled agent can cost tens of thousands of dollars in some cases, and losing experienced employees removes people who already know how to navigate complexity, Walters said.

Beyond technology

Reducing the burden on agents requires more than adding technology. Leaders should design tools around how decisions are made, ensuring they clarify what is happening, why it matters and what the agent should do next, Schmidt said.

Some companies are exploring dynamic routing that would give agents easier calls after emotionally difficult ones, but the model may be difficult to scale, McDougall said. Schmidt also pointed to breaks, workforce management design and task variation as ways to reduce strain.

Brown framed the problem more broadly: Companies cannot treat customer service as a role that can be boxed in and automated. The work is changing, and leaders need to involve agents in that change.

So, while AI may still reduce agent burden, it does so only when organizations design the work around context, workflow and human limits. Otherwise, the technology meant to make service easier can leave both customers and agents facing the hardest parts of the experience.

“There are no easy calls anymore,” Walters said.