92% of consumers escalate service issues from AI to humans within 10 minutes
More than half of consumers say they trust artificial intelligence (AI) tools to help resolve their customer service issues, as long as there is an option to escalate their question or complaint to a human employee, according to research published by Five9.
The San Ramon, Calif.-based provider of cloud-based contact center-as-a-service applications commissioned a survey of more than 3,000 consumers across the U.S., the U.K. and Germany to produce its 2026 Consumer CX report.
Five9 found that when no human option is available within a customer service interaction, trust in the technology drops to just 26 per cent.
Transparency throughout an interaction is also extremely important, the study found. For example, 71 per cent said they should know whether they are engaging with an AI agent or a human service rep. This was markedly higher among older consumers but still represented more than half of those in the Gen Z cohort.
Consumers may be growing accustomed to AI within customer service, given 54 per cent say interactions have shifted to AI, with less human contact. Not surprisingly, the top reason consumers say they escalated an issue to a service rep was due to an AI chatbot misunderstanding them, at 46 per cent. However nearly as many (42 per cent) said they simply prefer human service interactions.
“To get both CX leaders and consumers on the same page, contact centers must
evolve, and most importantly, bring consumers along with them,” the report’s authors wrote. “It’s not enough to simply deploy AI and hope for the best. The difference between success and failure will be in leaders’ execution with AI: how and where it’s used in workflows, the humantic data it’s trained on and the transparent policies and declarations of its use to consumers.”
360 Magazine Insight
Perhaps many consumers (in their professional lives?) have been hearting the repeated mantras about human-in-the-loop and how the best AI deployments augment human employees.
That could explain why a surprising number of people were okay with AI handling tasks like password resets, and not just managing their online orders, so long as reps were overseeing the technology.
Beyond the core stats, the report suggested brands prioritize fuelling AI platforms and applications with “humanic data,” or information about what reps do well, their most common mistakes and everyday processes. AI needs the equivalent of a job shadowing session, in other words, to meet consumers expectations in customer service.
It’s also worth noting that 29 per cent of consumers don’t believe AI handles their personal information in a secure manner. Communicating clearly what safeguards you have in place and how consumers can opt in or not could go a long way alleviating concerns and boosting AI acceptance.
This ungated 23-page report provides a lot more on the degree to which consumers find AI in customer service frustrating, including demographic differences (guess who tends to repeat themselves more often?), personalization trends and trust/transparency concerns.
