Customer service solutions generating AI, including the new AI agent category, can improve the service and, potentially, the whole customer experience (CX). But there is a disconnection between what AI can currently do and how agents of the contact center and customers use it.
A survey of agents in the contact center has revealed that if AI assistance is useful, AI tools do not fully understand the complexity of certain customer problems – and human agents still have to face the frustration of customers during interaction.
Another recent survey has revealed that the agents of the human contact center treat more difficult interactions of customers – both in terms of high complexity and emotions – because the tools fueled by AI manage relatively simpler information. These AI agents can both help and / or replace humans in some of the routine tasks they perform, in the contact center as well as in the largest knowledge workers.
Some studies have shown that if the tools fueled by AI can provide self-service options (that is to say, by “deviating” or “containing” these interactions with self-service options) and perhaps CX more broadly, other studies highlight a disconnection: customers do not always prefer their self-service costs, but many companies prefer that costs.
Others believe that even if the AI tools can indeed replace most, if not all, level 1 service / support, there will be a continuous demand – and perhaps increased – for the human key in more complex or nuanced customer -composition interactions.
Here are six stories that illustrate the above trends.
What the human agents of AI feel
An investigation into 152 Customer Service Agents In the United States and the United Kingdom, 24% of agents have classified customer frustration as their main concern. Twenty -one percent of the agents said that the AI tools they use cannot fully grasp the complex problems of customers and struggle with nuanced or complex requests. But 18% of the agents interviewed said that AI had saved them time and quickly find answers to customer questions.
Approach the agent’s professional exhaustion
While AI is deployed to take charge of customer routine requests, human agents eventually manage more complex and emotionally charged interactions. These more difficult calls increase so the cognitive load levels of agents and stress that 75% of the leaders of the North American contact center expressed themselves by their impact on the well-being of agents Digital CX survey in 2025 from Omdia.
Delegate to AI agents
According to Salesforce, AI AI AI agents that generate AI can be used alongside human agents to make them more productive. These are not “simple” chatbots or assistants fueled by customers that human agents encourage responses.
Instead, AI agents are given agency To question the databases and act, via APIs, to autonomously solve specific problems, such as product yields or appointments.
Show how Bots AI improve CX
With only around 60% of customers Ready to use voice / chatbotsCompanies must spend time building, training, designing and applying these robots to the problems they can manage. Companies must also educate customers on the reasons why robots are beneficial.
This could be resolved to present the success: a system message that says: “The bot solved your problem in two minutes. Depending on the current volumes, it would have taken 14 minutes.” This approach to present the value could help modify the desire of customers to use this option.
Humans and AI are the two CX keys
The vast majority (91.3%) of the companies questioned estimates that they offer excellent customer service on most interactions, according to a metrigy study at the end of 2024. On the other hand, a study from the beginning of January Metrigy revealed that only 14.7% of consumers considered customer service they had received as excellent. Much more – 40.6% – gave companies a “good” head sign, while just a little less than that – 36.3% – gave a “fair” note. Companies and consumers, however, agree on two fronts: AI, as well as competent and friendly human agents, are both Key for better cx.
Continuous need for human agents
AIA assistants and AI agents can indeed Help make human agents better, faster and more effective. These AI agents can completely replace human agents, at least for many common problems, leading to more and more self-service.
But sometimes human agents help customers diagnose their service / support problems before helping them. Even if level 1 support becomes almost entirely a self-service channel, level 2, level 3, level 4 customer service can persist because it is basically not the same thing as self-service-it looks more like human psychology.