Most consumers hate AI customer service chatbots. Here's why.
Nearly one in five consumers who used AI for customer support saw no benefit, according to Qualtrics' 2026 Customer Experience Trends Report. That failure rate - almost four times higher than for AI use in general - reveals a problem specific to how companies deploy the technology.
The issue isn't the AI itself. It's the incentives built into it.
Chatbots reflect corporate priorities, amplified
"AI doesn't change corporate incentives - it scales them," said Ben Wiener, global head of digital experience at Cognizant. If a company prioritizes minimizing refunds, reducing escalations to humans, or shortening call times, its AI system will optimize for those goals with mechanical consistency.
Human agents have always operated this way. But AI systems enforce these choices at higher volumes and without the judgment calls humans make.
Carmen Smith of California described the common experience: endless loops, repeated FAQ links, and information she'd already tried. "I'd rather speak to a human being," she said.
Consumers rank AI customer service among the worst applications for convenience, time savings, and usefulness. The problem, according to Isabelle Zdatny at Qualtrics, is straightforward: "Too many companies are deploying AI to cut costs, not solve problems, and customers can tell the difference."
The difference between deflection and resolution
There's a meaningful distinction. Deflection - refusing a legitimate request or trapping customers in loops - damages trust and business results. Enforcing actual policy rules is different.
"If the law says a refund is not available, a judge would adjudicate rather than argue continuously," said Terra Higginson, principal research director at Info-Tech Research Group. AI can enforce rules consistently without the strain of confrontation. But making legitimate refunds hard to secure is obstruction, not service.
In competitive markets, poor customer experiences spread quickly through social media and review forums.
Tom Eggemeier, CEO of Zendesk, said too many companies define "resolved" to include deflections and non-answers. His company only counts a resolution when the customer, business, and employee all agree the problem was actually solved.
What companies are doing differently
Some vendors are building AI systems around better incentives. Sierra, the conversational AI platform founded by former Salesforce co-CEO Bret Taylor, uses "outcomes-based pricing" - if the AI doesn't resolve the issue, it doesn't work for them either.
Decagon's CEO Jesse Zhang said he hasn't encountered a customer with deflection as an explicit goal. "People are very aggressive about optimizing for resolution," he said.
But guardrails matter. Some companies impose so many restrictions on AI judgment that the system can't actually resolve anything. Others deploy old-fashioned rule-based chatbots and call them AI, confusing customers about what they're actually using.
Escalation paths are non-negotiable
Every AI system needs a clear path to a human agent. Elderly customers, VIP accounts, and complex problems should have immediate escalation options.
Klarna, the fintech company, cut 40% of its workforce partly through AI implementation. But it later rehired some customer service workers after the AI struggled with complex tasks. The company's AI assistant now handles work equivalent to 800 agents with satisfaction scores on par with humans, according to a company statement.
At healthcare company NotifyMD, AI handles simple billing inquiries. But complex, emotional issues require humans. "There's no way for AI to bring the kind of understanding and empathy that a human being can bring, especially if the customer is upset or has a legitimate problem," said Jodi Miller, senior vice president of sales.
The scale ahead
AI chatbots may handle up to 80% of digital customer service interactions within five years, according to industry forecasts. Within three years, that figure could reach 50%.
The technology isn't going away. How companies choose to implement it - whether to cut costs or solve problems - will determine whether customers see AI as a tool or a barrier.
Learn more about AI for Customer Support and best practices for implementation.
Your membership also unlocks:
