AI call center is off to a rocky start

Artificial intelligence may be the future of customer service, but some early consumer reviews suggest you should be prepared to get frustrated, at least for now.

AI-powered chatbots can act as virtual concierges to direct customers to the right solution, but many customer service chatbots are still deflecting problems rather than solving them. Outright rejection of requests or sending customers into an AI-powered maze of uncertainty that leaves them too fed up to continue complaining is still common in the chatbot playbook.

“I hate AI customer service chatbots,” said Carmen Smith of Campo, California, who said she often falls into an endless loop when dealing with technology. “It seems like no matter what, they’ll all either direct you to some sort of FAQ list or just repeat information you’ve tried before and found missing,” Smith said. “I hate dealing with them, but unfortunately a lot of companies use them these days. I’d rather talk to a human.”

Smith is not alone. Almost one in five consumers According to the Qualtrics 2026 Customer Experience Trends Report, those using artificial intelligence for customer service saw no benefit from this experience. With a failure rate nearly four times higher than using AI in general, this figure points to something specific about customer service that makes it harder for AI to get it right. Consumers rank AI applications for customer service among the worst in terms of ease, time savings and usefulness. “Too many companies are using AI to reduce costs, not to solve problems, and customers can tell the difference,” said Isabelle Zdatny, head of thought leadership at Qualtrics XM Institute and author of the report.

There’s a simple business reason why many customers’ experiences are not positive. “AI isn’t changing corporate incentives — it’s scaling them,” said Ben Wiener, global head of Cognizant Moment, the digital experience practice of global technology and consulting firm Cognizant.

‘Continuously optimize’

Companies have always structured customer service around what they measure and reward. In many customer contact centers, human agents work within tightly scripted flows designed to limit discretion. In others, brands empower their employees to do whatever it takes to make customers happy.

“If leadership prioritizes minimizing chargebacks, reducing human escalation, or shortening call times, you can expect AI agents to reflect that philosophy into the experience — just as a human agent would. These were always business choices, and AI systems will implement those choices,” Wiener said. Wiener added that artificial intelligence can do this more consistently and in higher volumes. “AI will relentlessly optimize whatever metric it is given,” Wiener said. “Businesses need to be clear about what outcomes they want their AI systems to prioritize, because these systems will deliver exactly what they have been trained and measured to achieve,” he added.

“What makes them uncomfortable is automation that traps them in a loop,” said Shannon McKeen, professor of practice and managing director of the Center for Analytics Impact at Wake Forest University School of Business. Research on support automation shows that many conversations made with artificial intelligence are eventually transmitted to humans. But when systems fail to solve the problem or clearly explain a decision, customers often experience the AI ​​layer as an additional hurdle rather than a solution, McKeen said.

Diversion has advantages for people working in customer service.

The perversion of AI is justified when it is used to protect workers in jobs with high burnout rates and turnover rates, according to Terra Higginson, principal research director at Info-Tech Research Group. associated with mental health problems.

And in some cases, saying no is the right decision.

“If two people are arguing about a refund and the law says it’s not possible, the judge will make the decision instead of constantly arguing back and forth. This often happens in scenarios where the agent is to the unhappy customer,” Higginson said. “This streamlines the process of enforcing rules and regulations rather than making refunds more difficult,” Higginson said, adding that AI can consistently enforce rules in a way humans cannot, “without the arguing and back-and-forth tension that comes with being yelled at for following company rules.”

On the other hand, making it difficult to secure legitimate refunds is, and always has been, bad business. “This is not a service, it’s a disability,” Higginson said. Higginson added that this is a particularly bad business model in a competitive market where digital ideas can spread quickly through forums and social media.

Consumer-facing chatbots are here to stay

Zendesk CEO Tom Eggemeier says many companies define “resolved” interactions as including deviations and non-responses. Zendesk only counts the solution if the customer, business, and employee agree that the problem has actually been resolved. “Artificial intelligence is a tool, not an end,” Eggemeier said.

One solution he thinks will likely happen in the not-too-distant future is for consumers to have a personal AI representative to handle company chatbots, allowing AIs to solve low-level problems.

Consumers may need help.

Eggemeier predicts that within three years, 50 percent of digital customer service interactions will be managed by AI, and that percentage will rise to 80 percent within five years.

Jesse Zhang, CEO of customer service chatbot creator Decagon, which tripled its valuation to $4.5 billion in 2025 after signing more than 100 enterprise deals in consumer-facing industries, says companies that aim to divert customers will lose money in the long run.

“We haven’t come across a single customer who intends to refer,” Zhang said. “People are very aggressive about optimizing for the solution,” he added.

Sierra, the interactive AI platform founded in 2023 by former Salesforce CEO Bret Taylor and former Google executive Clay Bavor, says its business model uses “outcome-based pricing” and thinks it’s an important way to approach these new interactions. “If we can’t fix the problem, if it doesn’t work for customers, then it doesn’t work for us,” a spokesperson for Sierra said.

Zhang acknowledged that there can be subjectivity around the issue on the client side, and one person’s decision may be a bias for another. But he said it’s a company’s job to have AI smart enough to make decisions. “You can’t say no to everything, you can’t say yes to everything. You want to find a solution,” Zhang said.

What should never happen is not a dead end, but an “escalation path” for customers who don’t get what they need from AI answers.

There are situations where a clear and fast path to a human agent must always be available, such as for elderly customers, VIP customers or particularly complex issues.

A widely cited example of AI chatbot implementation occurred at fintech Klarna, where AI played a significant, if not sole, factor in the recent 40% reduction in headcount. However, the policy change to prioritize artificial intelligence ultimately resulted in the company making the following decision: rehire some Customer service workers after AI technology showed lower quality performance on some more complex tasks. Klarna remains committed to its use of AI, launching an AI assistant that did the work of 700 customer service agents at launch and has now expanded to 800 agents, a company spokesperson recently told CNBC. The spokesperson said the AI ​​assistant is now receiving more customer requests and customer satisfaction scores are on par with human agents.

The market is evolving rapidly, which can lead to a variety of customer experiences. “Sometimes consumers don’t know the difference between an old-fashioned chatbot and AI. Old-fashioned chatbots can’t do things and solve problems,” said a Sierra spokesperson. There are also some brands that are being “super cautious” about introducing new AI chatbot technology and putting so many guardrails on their AI reps that the judgment needed to fix the issues isn’t even there. “They do it because they’re worried they’ll make mistakes, but the guardrails need to be reasonable,” he said.

As in healthcare, there are more complex AI use cases on a sector-by-sector basis. At NotifyMD, AI was used in customer service to tackle simpler problems, such as responding to billing-related customer calls. But Jodi Miller, senior vice president of sales, said people still matter in anything complex and emotional. “There is no way AI can bring the understanding and empathy to the table that a human can, especially if the customer is upset or has a legitimate issue,” Miller said. “I think the key going forward for all of these companies will be to be very thoughtful about the use of AI and to make sure that AI helps rather than hinders the people who need help the most,” Miller added.

Zhang believes that the future of customer service will be AI and that the AI ​​agent will have memory and be able to handle all types of customer service queries. “At a very high level, every business will have AI at the front end with a single unified agent across all channels,” he said.