Source: Pexels
It’s possible to use AI agents in customer-facing roles without sacrificing the empathy that customers expect. I sometimes see leaders hesitate to deploy agents for customer interactions because of legitimate concerns that AI agents can’t offer empathy. However, with the right implementation and monitoring strategies, customer-facing agentic AI can make interactions more efficient and more empathetic by facilitating calls and resolutions. Achieving this kind of AI implementation can help set your brand apart at a time when consumers are actively seeking empathetic experiences.
Empathy is now a leading CX differentiator
The importance of empathy in customer experience is now well-documented. A 2025 global study by Zurich Insurance, YouGov, and the director of Stanford University’s Social Neuroscience Laboratory showed that consumers place a high value on empathetic experiences with brands and their representatives. Among the findings that matter for CX:
60% of customers only do business with “companies that demonstrate genuine care”43% have stopped engaging with a brand because they felt it lacked empathy71% of consumers don’t think AI can replicate human connection.
Critically, 61% of consumers are willing to pay more to do business with brands that show empathy. So, if companies that deliver empathy can retain more customers, and those customers are willing to spend more with them, then empathy is a quantifiably valuable brand asset.
But if those same customers don’t think AI can be empathetic, how are you supposed to use AI to improve empathy? This is an urgent question because more than 90% of customer service leaders feel they need to deploy AI this year. The answer is to implement it in a way that enhances empathy rather than undermining it.
Using AI agents to support more empathetic CX
Customer-facing AI agents don’t need to match human levels of empathy. They just need to handle high-volume, straightforward tasks that don’t require deep human interaction so human agents can focus on the customer cases that require empathy and time to resolve. For example, chatbots can save customers time by helping with simple problems such as
Resetting passwordsScheduling appointmentsProviding order status and delivery updatesTroubleshooting basic issuesSurfacing information in FAQs.
By handling these interactions, AI agents can reduce wait times for customers who want to resolve easy issues fast. This ability to solve simple problems fast is one reason Gartner expects agentic AI to handle 80% of common customer service issues autonomously by 2029.
AI agents can also route customers with more complex problems to the appropriate human agent. Those agents can then draw on their emotional intelligence and judgment to resolve more complicated issues with the right tone for the customer’s needs. Agentic AI that helps human agents resolve customer issues more effectively also gives employees a greater sense of accomplishment. That, in turn, can boost employee confidence and empathy in future customer interactions. This is where empathy becomes an asset that strengthens customer relationships. Because empathy and discernment are so important in these cases, Gartner estimates that half the companies that initially planned to cut their customer service staff in favor of AI will rescind those plans by 2027.
The ideal result is a hybrid model that integrates customer-facing AI agents and human agents to meet the needs of each customer, whether they’re interested in handling a basic task fast or need to talk to a person about a larger issue. AI agents can also play internal roles in this kind of customer service model. Monitoring customer interactions across channels at scale can improve the quality of human and agentic service by ensuring that interactions comply with brand guidelines and relevant regulations. AI monitoring agents can also quickly deliver training updates for new hires and for existing employees who need additional support.
For example, one of my organization’s telecom clients has built a more empathetic customer support experience by replacing manual QA with agentic AI-powered QA that monitors and analyzes all customer interactions in real time. This allows the AI agents to estimate CSAT and NPS scores on interactions in progress so supervisors can coach agents during interactions for better customer outcomes. Supervisors also get alerts about critical calls so they can step in as needed. After implementing agentic AI for QA, the telecom realized a 41% increase in NPS.
Plan a controlled AI agent rollout
Because AI agents have the potential to resolve so many customer issues so quickly, you might be tempted to roll out multiple use cases at once. This approach can cause problems that undermine internal support for AI in CX and alienate customers. It’s wiser to start with one high-volume, low-risk use case, monitor and learn from the deployment in real time, and only scale up and add new use cases after the first one is working well.
For example, Airbnb’s custom-built AI customer service agent launched in Q1 2025 with a limited rollout to half of US users. That initial deployment reduced contacts with live agents by 15%. After that initial period, the company gave all US users access to the agent and then gradually expanded its reach. By February of this year, AI agents were taking care of one-third of customer support contacts in the US and Canada.
Tracking ROI on empathetic customer service
As you plan your customer service department’s agentic AI use cases, select KPIs that will allow you to track progress and return on your agentic investment. ROI-related KPIs to consider include time to resolution, number of contacts handled by AI, CSAT and NPS ratings, customer churn, and employee turnover. A well-designed hybrid system that uses AI for the basics and frees humans to provide empathy and discernment when it’s needed can retain customers and protect employees from burnout. As customer churn decreases, average customer acquisition costs may decline while CLTV increases. And when employee turnover slows, there’s more room in contact center budgets for new technology and training to improve CX even more.
So, although AI agents aren’t capable of showing empathy, they can create the conditions that unlock it. Finding the right roles for AI and human agents can create the experiences that customers say they want now, and which they say are so hard to find.