Be wary when dealing with your bank’s chatbot

Financial institutions are increasingly using chatbots, which put out human-like responses using computer programming, to communicate with customers to save money.

Chatbots sometimes have human names and use popup features to encourage interaction. Some chatbots use more complex technologies such as “artificial intelligence” to respond to customers.

The Consumer Financial Protection Bureau has received hundreds of complaints from frustrated customers trying to receive timely, straightforward answers from their financial institutions or to raise a concern or dispute.

“To reduce costs, many financial institutions are integrating artificial intelligence technologies to steer people toward chatbots,” said CFPB Director Rohit Chopra. “A poorly deployed chatbot can lead to customer frustration, reduced trust, and even violations of the law.”

About 37 percent of the U.S. population interacted with a bank’s chatbot in 2022, a number that will grow.

Among the top 10 commercial banks in the country, all use chatbots to communicate with customers.

Financial institutions say their chatbots offer a variety of features to consumers such as retrieving account balances, looking up recent transactions, and paying bills. Much of the industry uses simple chatbots with keywords or emojis that trigger preset, limited responses or route customers to a Frequently Asked Questions section or page.

Other institutions have built chatbots by training algorithms with real customer conversations and chat logs, such as Capital One’s Eno and Bank of America’s Erica. Also, the banking industry has begun adopting advanced technologies, such as chatbots that can generate an answer.

In a study, the CFPB found chatbots pose risks including:

• Noncompliance with federal consumer financial protection laws. The information chatbots provide may not be accurate. The technology may fail to recognize that consumers are invoking their federal rights, or it may fail to protect their privacy and data.
• Diminished customer service and trust. Consumers’ problems could be dire and urgent. They may face repetitive loops of unhelpful jargon and struggle to get the response they need, including not being able to talk to a human.
• Harm to consumers. Consumer may select the wrong product or service. They could be assessed fees or other penalties if they receive inaccurate information on making payments.

People who are having issues getting answers to their questions due to a lack of human interaction can submit a consumer complaint by going to the CFPB’s website or by calling 855-411-2372.

Comments

[azure]

All the chatbots I've tried to use (not on bank or credit union sites) have been useless. If I didn't enter a question or phrase within the chat bot's "vocabulary" it either provided links to "help" that weren't even close to the "help" I sought or just kept asking me what I needed help with. As far as I can tell, many corporations want to be able to say they're offering customer assistance but spend as little money as possible providing it, and if your problem isn't solved by a FAQ, that's ok with the corporation, must be since most make it increasingly difficult to speak to live and well trained (and adequately compensated) person, whether online or by telephone.

Good luck w/getting through to a live person who actually has the knowledge/skills/authority to help you. Consumer Cellular whose on hold announcement keeps informing callers that it won a JD Power award for great customer service (probably in 2000) now forces customers who call (who knows what their website is like) to go through a 2 level automated voice menu system (w/mediocre word recognition software), only then can you--maybe-- speak to a live person AND you now MUST HAVE a 4 digit pin before anyone will talk to you. Since I rarely speak or deal w/CC other then paying the bill & became a customer before pins were required, I didn't have one. It took me a few minutes to "persuade" the automated menu system to let me speak to someone anyway, I had to say that I'd "forgotten" my pin. The machine couldn't deal w/my saying I simply didn't have one, had never had to have one prior to that call. It was clear that unless I lied & said I forgotten my pin, the call would be disconnected. EvenWhen I reached a live person, I was given incorrect information that didn't solve a problem for which I'd requested assistance. I had to call again.

The pin is supposedly for 'security' but a 4 digit pin isn't a secure password. I suspect it's just all CC was willing to pay for, the tech was purchased years ago & has never been udpated or upgraded. There's data security breaches happening frequently it seems and they'll keep happening until our Congress and all state legislatures start imposing substantial fines and allowing consumer lawsuits w/attorney fees for the consumer's laws paid for by the business. That seems unlikely, the opposite, making it even more difficult to hold corporations responsible for not keeping confidential data secure, seems far more likely. Right now, there's really no real (to a large corporation) penalty for not providing adequate security for the confidential information they require.
The EU seems far more willing to act to protect consumer data privacy, at least for now. Be interesting to find out if there are fewer chatbots used by EU businesses or if they're being widely adopted there too.

I wonder how much "chatbots" were able to help all the customers calling Chase re: the double billing Zelle did? Apparently it's not the first time Zelle's caused problems. This time, people were seeing bills deducted twice from their checking accounts by Zelle, leaving some w/out money to buy groceries. Chase claims to have apologized/fixed the problem (& credited accounts) rapidly, I wonder if Chase will also pay any insufficient funds fine/fees caused by Zelle's screw up? Probably US consumer and banking regs don't require it, even if they should. But you're more likely to know if banks do or don't have to pay for such charges when the bank or an app provided by the bank causes those charges then I am.

[Carol Cassara]

It is so incredibly frustrating to experience the distance companies put between themselves and customers.

[Rita]

With AI advancing, I think the chatbot problem is going to get worse.

[Jennifer]

I'm glad my bank has not stooped to using AI yet. Yes, we do have an automated system where clients can listen to their transactions and conduct business if they want but there is always an option to get out and speak to a human for assistance as well. Many of our older clients trust phone banking but don't trust online or digital banking.

[Rita]

I belong to two credit unions. I can always talk to a person without waiting too long. I have tried to use chats on some websites when I have questions. They don't seem to work very well. They have had names, so I thought a person was communicating with me in a chat. However, they could have been chatbots. Maybe that's why they didn't work very well.

[Diane]

I've never interacted with one successfully. 100% of the time, it's useless!

[Rita]

Same here. I thought it was a person, since when I was doing the chat, the "person" had a name.

[Rebecca Forstadt Olkowski]

I got a question answered by a chatbot on Capital One when it told me where something was that could have easily been found if they put it on the home page. Some chatbots have been helpful but others are useless. They never understand the question.