TECH on Deck: The AI Perception Gap — You’re Solving Last Year’s Problem

Enterprise CX leaders are operating on an outdated picture of what AI can do — and that gap between perception and reality is showing up in roadmaps, vendor evaluations, and success metrics. The capability frontier has moved so fast that assumptions from even six months ago are now a liability. Teams are asking last year's questions, vendors are getting away with last year's demos, and roadmaps are being set based on last year's reality. George Humphrey, SVP of Research at TSIA, and Merlin Bise, CTO of Inbenta AI, will show exactly how far the gap has grown, why the old evaluation frameworks are actively producing wrong answers, and what it costs to keep using them. If your team's mental model of AI hasn't been stress-tested in the last 90 days, consider this session your reset.

AI in CX is advancing faster than the mental models enterprise leaders are using to evaluate it. Teams are still asking “can AI handle our FAQ volume?” when the real question is “can AI complete a multi-step transaction on the customer’s behalf and prove it did so correctly?” The loudest opinions in most evaluation meetings are based on vendor demos from 12 months ago, and those opinions are setting roadmaps that are already behind. Deployment timelines, success metrics, even the definition of first-contact resolution are all shifting underneath CX leaders who haven’t updated their assumptions. It’s a perception gap that’s compounding every quarter.