Your customer service representative would like to speak to a human

Customer service representatives often encounter AI agents on the other end of the line, questioning whether they are speaking with a human or an artificial intelligence. These AI agents are increasingly being used to call customer service lines and negotiate bills or cancel subscriptions on behalf of consumers. While some customers find AI deployment helpful, it presents a new challenge for human agents who must now deal with customers that are actually AI.

The growing use of AI agents in call centers has raised questions about how to determine if an agent is human or AI, how to engage with them, and whether they are authorized to act on someone's behalf. One insurance company agent who works in claims says she gets calls from the same AI agent every shift, who asks a series of questions about a car accident claim. The agent says the AI gets irritated when it doesn't have the answers and insists she wait on hold for an associate who never arrives.

Another call center agent in the customer retention department of a major credit card company says AI callers are often on a fact-finding mission, trying to map out the system. She says she can usually tell the AI is not a human customer because it lacks enthusiasm and hesitation in its voice.

The quality assurance team at TTEC, a customer service outsourcing company, began noticing something off about AI callers about a year and a half ago. Some callers simply sounded weird, and once they realized they were AI, TTEC asked clients how they wanted to manage these calls. Generally, the response was to take the call, but proceed with caution.

The AI often doesn't have enough information to complete necessary security measures or prove it's not a fraudster. Companies must weigh the risks of security and data when dealing with AI agents, and the challenge of determining whether a call is from an actual customer, a convenience, or a security risk.