Key Takeaways
- Capacity’s Agentic Consumer Report finds that 35% of consumers have used an AI agent to handle a customer service request
- Consumers who use AI ask more specific questions, challenge answers and request policy explanations
A report from agentic support automation platform Capacity dives into ways consumers using their own AI tools and agents during customer service interactions. as businesses deploy AI to improve the customer experience
The Agentic Consumer Report revealed that consumers, especially younger generations, are turning to personal AI tools more often during customer service interactions. These tools are creating more empowered customers who feel increasingly comfortable pushing back on the responses they receive from customer service representatives and challenging their knowledge of company policy, according to the authors of the report.
The reported noted while 35% of consumers surveyed have used an AI agent to handle a customer service request at least once, that figure rises to 49% among Gen Z and Millennial respondents.
How Should Businesses Measure Customer Service Success When Customers Use AI?
For years, customer service leaders have used AI to deliver faster answers and higher resolution rates, said David Karandish, founder and CEO of Capacity.
“But as customers begin using the same technology to challenge those answers, the goal has to shift,” said Karandish in a statement. “Businesses need to rethink how they measure success, moving beyond speed and efficiency to consider whether their answers are accurate, their decisions are defensible and their customers can trust the outcome.”
How Is AI Changing the Way Consumers Interact With Customer Service?
The findings of the report highlight several ways AI tools are becoming a part of a consumer’s customer service experience. Among respondents, 34% said AI makes them ask more specific questions during customer service interactions, and 23% said they would ask representatives to explain or cite company policies because of AI. Other results from the report include
- AI is introducing incorrect information into conversations: 55% of respondents have used an AI tool, such as an agent or LLM, to assist with a customer service request, but 44% said AI had contradicted a company representative or its policies.
- Consumers trust AI as much as a company representative: If AI were to give a consumer information that conflicted with what a customer representative had said, consumers are equally likely to trust either one.
- The escalation path is shifting: 20% of consumers said they would escalate issues more quickly when using AI tools, while 21% would trust their AI agent to connect them with a human representative, and then join the conversation.
How Can Contact Centers Prepare for AI-Assisted Customers?
Human representatives and AI agents must be prepared to support informed and assertive customers. The report highlights consumers using AI are asking more specific questions, challenging answers and requesting policy explanations. Contact centers will need to equip representatives with accurate company information so they can address discrepancies and stand behind their decisions and responses with confidence.
Businesses must prepare for interactions between customer-owned and company-deployed AI agents as consumers are using AI to reach human representatives and feel more confident escalating issues when assisted by AI. Capacity officials note that contact centers will need clear escalation processes and safeguards against repetitive AI-to-AI exchanges, ensuring routine requests remain automated while human representatives focus on complex issues.
“CX strategies need to adapt as consumer behavior shifts,” said Karandish. “The best way to do so is with a continuous learning loop that recognizes when consumers are more likely to challenge an answer. Insights from those scenarios will be crucial to preparing both AI agents and human representatives for more complex, AI-assisted interactions.”
The Agentic Consumer Report is available by clicking here.
