AI ischanging the way contact centers work, bringing bothjob reductionsand new responsibilitiesfor existing employees. This trend also brings unprecedented challenges to customer service leaders.

"Let's not confuse concepts," said Neville Letzerich, Chief Marketing Officer of Talkdesk, during a panel discussion at Customer Contact Week in Las Vegas last month. "This is not a change in operating model, but a paradigm shift. If you think 'just putting AI and people together will solve everything,' you're mistaken."

Letzerich pointed out that employees generally feel threatened, worrying that AI will replace their jobs. Leaders need to carefully consider how to introduce and scale AI in call centers while properly addressing employee concerns.

Experts say an effective strategy is to keep employees informed in a timely manner about possible layoffs or new job opportunities. Leaders should avoid viewing AI as a black-and-white choice—in some service journeys, having AI handle initial interactions followed by human intervention may be the optimal solution.

Nicole Kyle, Managing Director and Co-Founder of CMP Research, believes that regardless of how plans proceed, every leader must consider the impact of AI initiatives on the workforce. Addressing employee skepticism and maintaining high frontline morale are crucial for sustained success.

"Executives are under time pressure," Kyle told CX Dive. "How do they balance the transformation of the CX technology roadmap with people-related challenges? The best CXOs and customer contact leaders understand that both must be addressed."

Be transparent with employees from the start

Jessica Gupta, Chief Operating Officer of data technology company InfoPay, believes that whether jobs are lost or new opportunities emerge, companies must be transparent with employees about how AI will impact their call center operations.

Gupta said InfoPay began informing employees about AI plans about two years before determining whether layoffs would occur. Despite this, its communications clearly acknowledged that jobs could be affected.

"I was very frank that I wouldn't pretend this wouldn't change the staffing we need, nor would I pretend it wouldn't change how we work," Gupta said. "But we'll go through it together, proceed carefully, and ensure fairness. We are very clear-headed because we want people to have time to prepare."

According to Gupta, InfoPay took a very systematic approach. The company not only helped teams prepare for possible layoffs but also began discussions with employees identified to stay after AI arrives about what new roles might look like.

For example, customer service agents with deep knowledge of certain topics joined analytics projects to help launch AI initiatives. They participated early as experts, focusing on the most critical aspects of the support experience. As InfoPay added AI capabilities, these employees continued to be involved in development and testing efforts.

"We did have to make layoffs—cost savings are real—but we were able to build teams that are now doing exciting things," Gupta said. "We created careers for people that didn't exist before."

Gupta believes one major benefit of this approach is that it helps maintain high morale as AI takes on a larger role in contact center operations. Agents know work is changing, but they feel a greater sense of control over the technology driving that change.

"I think involving them in the process reduces fear," Gupta said. "They will go through this journey with us and grow together."

Remember the most suitable role for humans

Bob Sacunas, Vice President and Strategic Industry Executive at AI solutions provider UiPath, said technology is not automatically the answer. In scenarios requiring judgment, humans remain the ideal choice.

Jonathan Rosenberg, Chief Technology Officer of contact center provider Five9, noted that in certain industries, having customers interact with AI first before being transferred to a human is not a replacement but part of standard operating procedure.

"Self-service is not a black-and-white matter," Rosenberg said during the panel discussion. "Many interesting use cases actually involve AI systems performing initial data collection, triage, and then purposefully transferring calls—not transferring because of failure."

Rosenberg cited loan applications as an example, where AI can efficiently collect necessary customer information and then transfer the call to a human loan specialist who has the judgment and empathy to put the customer at ease.

"If you are customer-centric and put them first, it doesn't mean they no longer talk to people," Rosenberg said. "You can achieve cost savings, deliver excellent customer experience, and ensure your agents still play a role in important interactions."

Sacunas believes older forms of AI also have their use cases. Agentic AI is most suitable when a degree of reasoning is needed; but for simple, repeatable tasks, simple deterministic AI is often faster, cheaper, and more reliable.

"The way to avoid false promises is to ensure you choose the best tool for the job, rather than over-relying on AI and trying to apply it to everything," Sacunas said.