34% of consumers think AI-powered support makes things harder
Artificial intelligence (AI) may help speed up certain service and support processes, but that doesn’t mean they’re delivering the outcomes customers want and expect, according to research published by Glance.
The Wakefield, Mass.-based provider of software that enables “cobrowsing” and other guided customer experiences fielded a survey of more than 6,000 people in the U.S. to produce its 2026 CX Trends Report.
Glace found that while speed is generally considered a good thing in customer service, 75 per cent of those surveyed said AI had produced a fast response that left them frustrated. This includes being routed by a chatbot to human reps for basic questions, or providing stock answers that fail to address an issue.
Perhaps not surprisingly, the lack of effectiveness in AI-powered service lead 87 per cent to say they would be unlikely to remain loyal to a company that removed human support entirely.
On the plus side, 44 per cent said they always try self-service options first, suggesting that well-designed AI systems could become more viable over time.
“The next phase of CX won’t be about doubling down on speed or pushing more customers through automated funnels,” the report’s authors wrote. “It will be about building AI that’s anchored in real customer intent, clean data, and human-centered design.”
360 Magazine Insight
Glance’s survey appears to have been limited to a fairly small set of questions, and much of what you’ll find in this gated 24-page report is culled from third-party research from Gartner and other firms.
That said, the responses Glance gathered included some free-form answers that perfectly encapsulate why some consumers are frustrated with AI. This one stood out:
“When I ask the AI for a human representative, I want a human representative. I don’t want to keep asking over and over again for one, I want one the moment that I ask for them. Because obviously my problem is too difficult, or not being understood by the AI.”
After candidly using the data to show AI’s shortcomings, the balance of Glance’s report is much more prescriptive, drawing upon case studies and insight from its subject matter experts on designing thoughtful, more human-centered service and support experiences.
Instead of using AI primarily for digital deflection, for instance, Glance suggests building “digital ridealongs,” where technology provides genuinely helpful accompaniment amid self-service experiences.
One other important stat from Glance’s own data includes the fact that only seven per cent say they don’t often have to repeat themselves when switching from one channel to another. As we move into 2026, probably the biggest switching we’ll see is between AI and human beings, and that’s where synchronizing customer data will become the most critical. Glance calls this “context continuity.”
Overall, this is a worthwhile read for any CX leaders who want to reflect on the state of AI automation and rethinking their brand’s approach over the next year and beyond.
