AI and automation are transforming contact center solutions, making customer service one of the most common ways that people will interact with AI in their daily lives.
The advantages of well-executed, advanced contact center solutions are undeniable: reduced queue times, less friction in issue resolution, and more consistent service across channels. As such, technology adoption is very high—90% or more of businesses with customer experience (CX) functions use these solutions. However, usage tends to be inconsistent, with varying levels of sophistication, modernity, and integration of advanced tools across industries and organizations.
One major variable is the human element. Often, businesses require a hybrid of technology and agents to address the entire CX breadth. And agent skillsets, training and overall comfort with different CX tools can vary, potentially impacting operations and outcomes.
In ongoing research on this topic, I’ve come across many excellent studies that center on CX decision-makers, which is logical given their role in selecting and deploying technology, and visibility on ROI and outcomes. However, I thought it might be interesting to hear from the agents using these tools daily, especially considering how their engagement can impact CX success.
In February 2025, I conducted a brief survey of 152 customer service agents in the US and UK to understand how these tools function in practice and where challenges remain. While the survey’s size and scope didn’t allow for deep dives into specifics such as vendors or industries, the emerging themes provided a reasonable indication of AI’s impact on agents and suggested areas for improvement.
The most common challenge—and perhaps not a surprise to anyone who’s ever been on hold with a less-than-advanced center—is customer frustration with automated responses. Nearly one-fourth (24%) of agents ranked customer frustration as their top concern. On a similar note, 21% of agents noted that these tools cannot fully grasp complex customer issues, and struggle with nuanced or complex inquiries.
Reliability is another concern: 20% of agents said they frequently double-check AI-generated suggestions, reducing the technology’s intended efficiency. And 13% cited system integration issues, which can disrupt workflow. A 2025 survey of CX decision makers conducted by Nextiva (The 2025 CX Landscape | Nextiva) reinforces these findings, with 90% of those respondents citing “friction” coming most often from integration issues, legacy technology, or employee resistance.
Despite these challenges, agents in my survey still pointed to efficiency as the main benefit of AI-powered tools. The most cited advantage—selected by 18% of respondents—was the ability to save time and quickly find answers. Other key benefits included AI’s ability to summarize customer issues, handle simpler tasks, and suggest helpful responses. At the bottom of the list were detecting customer sentiment and synthesizing omnichannel communication, though limited adoption and availability of these tools is likely what lead to the lower rankings.
In fact, sentiment analysis had the widest gap between availability and demand—more than three times as many agents wanted the tool as those who currently had access to it.
Finding challenges in a user’s experience does not necessarily mean a technology is ineffective: agent experiences depend on factors such as training, workflow design, and the specific AI tools in place. In other words, frustrations may stem from how these tools are implemented rather than their capabilities. Conducting more in-depth user research could pinpoint areas where agents can be trained to maximize AI’s potential.
All the same, the data underscores the need to evaluate automation holistically. As the human component remains central to the equation, training, optimizing system integration, and ensuring agents have the flexibility to adjust AI-driven processes when needed should be at the top of any contact center’s strategy.
Automation is a tool, and its success depends on how well it is implemented and integrated. Decision maker-based research offers valuable insight into strategy, adoption, metrics, and outcomes. Agent perspectives could add depth by showing how these tools perform in practice, and help reveal what would improve a contact center’s usability, efficiency, and overall impact.