AI chatbots leave Korean consumers angry and frustrated

Touted by service workers as a revolutionary tool, the early stages introduction of generative AI services as frontline customer service providers has led to Korean customers expressing disdain for the already divisive technology.

KB Kookmin Bank runs a weekly ″KB Senior Lounge″ service in which staff members visit community centers for the elderly to provide them with banking services. [KB KOOKMIN BANK]

A nursing aid in her 50s, Shin just wanted to get some banking done.

Her attempts at talking with a bank’s chatbot had been going nowhere, which she said felt like a “prolonged automated response system (ARS).”

Commands to the chatbot asking for information on withdrawing from a fixed deposit were made as clear and concise as possible, but to no avail. Once again, the conversation with the chatbot met a dead end after it redirected her to unnecessary services.

“I reached out to a call center in the first place because I hadn’t been able to find the information I specifically wanted online. For the system to redirect me to the home page seemed useless,” said Shin, who had been hoping for a "short conversation" with a human agent.

Touted as a revolutionary tool by businesses, the early stages introduction of generative AI services as front line customer service providers has led to Korean customers expressing disdain for the already divisive technology.

Salesforce's July survey showed 39 percent of customers found chatbots challenging to use. [YOO YOUNG-RAE]

It’s not just Shin. Poorly integrated chatbots were the most commonly reported challenge among digital financial services by a sizable margin, with 39 percent of survey participants finding them taxing, according to California-based software solution company Salesforce.

The survey, which was published in July last year, asked 6,058 customers in 12 countries across five continents about the challenges and expectations they have in their digital financial experiences.

Automated chatbots, often advertised by banks as AI chatbots, were initially introduced as a measure to process large volumes of transactions simultaneously, by providing quick automated answers to simple commands. They were expected to reduce congestion for employees during busy hours and provide banking services to customers who were calling outside of official working hours.

Major commercial banks including NH Nonghyup, Shinhan, Woori, Hana and KB Kookmin Bank have all incorporated chatbots powered with generative AI in their customer service, either for use by employees to increase productivity through quicker searches in the bank’s database and to fill out paperwork, or for direct customer interaction. KB Kookmin Bank made headlines in December last year after it notified 240 call center employees of their imminent layoff, citing the high volume of AI use as the reason for the center’s closure. The workers’ employment succession was confirmed in January, but people’s criticism did not end, especially as they thought chatbots have not been able to perform at a level to replace human input, and fewer employees meant longer waiting times to get hold of a worker.

It was on one such viral post criticizing AI chatbots, updated on Jan. 20 on X, formerly Twitter, that online users piled on to express their frustration at a service that is increasingly becoming the norm.

“K bank fired 240 call service workers while citing the increase in AI use, which ended in them being redeployed across the business. I mean, [the replacement would be justified] if our country’s AI chatbots were at least at the level of ChatGPT or Claude,” the original post read.

The user then recalled a time when they waited 40 minutes to get hold of a human agent. A different user claimed they wasted the same amount of time using Hana Bank’s AI ASR, but resolved the issue in five minutes with a human worker.

Most users said their queries were only answered through human input. “Not once have bank chatbots been able to solve my problems, so I’m always at a loss looking for a person. Chatbot consultation is 100 percent a waste of time,” a user said.

Others complained how banks now immediately suggested customers use AI chatbots while creating more hurdles to connect to a human agent, and suggested that the banks’ claims of an increase in AI usage were inflated.

While even the somewhat tech-savvy X users struggle with the service, the older population faces obvious disadvantages when dealing with chatbots, which primarily direct customers to online information on mobile devices.

To deal with a growing disparity in the accessibility of its services, major banks have bolstered efforts to offer customized services to senior citizens who find the digital transition difficult, such as KB Kookmin Bank, which runs a weekly “KB Senior Lounge” education service in which a van visits community centers for the elderly to providing banking services.

While banks serve as a prime example, it is not the only industry where the integration of AI-powered customer service chatbots has left customers dissatisfied.

The Korea Management Association Consulting (KMAC) lowered the score in its evaluation of service quality and empathy levels provided by call centers in its 2023 service quality index report. The evaluation was made on 334 call centers in companies over 55 industries, of which the travel and airline industries’ call centers especially dragged down the average quality of service.

The organization cited two main reasons for the drop in customer satisfaction. The first was the longer waiting times for customers due to an insufficient number of available consultants. The second was that inquiries could not be resolved by chatbots and phone bots, leading to a higher concentration of customers demanding services from human agents.

This indicates a need for companies’ honest review of the level of technology and customer satisfaction in their transition toward AI.

“The level of quality of call center services was expected to become more advanced but instead has been stagnating or declining due to rapid changes in their operations and technological environments,” Lee Ki-dong, director of KMAC’s business value evaluation division, said. “I believe we are at an important turning point.”

BY KIM JU-YEON [kim.juyeon2@joongang.co.kr]