AI rollouts may be outrunning users’ trust, executives say
Two EY leaders say companies risk creating AI systems that work quickly but leave customers and employees unsure about decisions.
By Maya Lindqvist · Senior Technology Correspondent
3 min read
Companies pushing artificial intelligence into customer service, finance, health care and internal workflows may be overlooking a basic design problem: people often need more time than machines to understand what is happening. Patricia Camden and John Dubois of EY say that mismatch can weaken trust even when an AI system performs as intended.
Camden, EY Studio+ customer experience and loyalty leader, and Dubois, EY Americas AI strategy leader, describe the issue as a “tempo gap.” They use the term for moments when automated systems move faster than users can follow, assess or accept with confidence.
According to Camden and Dubois, many organizations still frame AI adoption around automation, productivity and speed. In their view, that framing misses how faster workflows can increase the mental burden on customers and employees who must judge AI-generated recommendations, decisions or prefilled information.
The EY leaders said earlier digital tools generally operated at a pace people controlled. A user searched, filled out a form or advanced through a process, and the software responded. AI changes that pattern, they said, because systems can infer intent, recommend actions and advance an interaction before a person has fully reviewed the situation.
They pointed to several examples from their work with enterprise clients. A traveler may be rebooked after a flight cancellation before comparing alternatives. A financial customer may move through an application quickly enough to approve significant terms without fully absorbing them. A patient completing online medical paperwork may see sensitive details filled in automatically before understanding how the information will be used.
Camden and Dubois said those cases do not necessarily show technical failures. The systems may be doing what they were built to do, they said, while still leaving users uneasy enough to recheck information or pause before continuing.
The concern is sharper in settings involving money, health care or sensitive personal data, according to the EY executives. They said people may welcome quicker service while still wanting to know why a recommendation appeared, what assumptions shaped it and when a human should intervene.
EY’s leaders said they already see effects inside organizations adopting AI. In some cases, teams spend more time verifying AI output they once would have trusted. In others, workflows designed to save time slow again as manual review returns to the process.
Camden and Dubois argue that the next stage of enterprise AI adoption will depend on matching system speed to human understanding. That may mean speeding up routine decisions, adding pauses before approvals, showing uncertainty more clearly or recognizing when a customer needs reassurance rather than another automated response.
They called this “intentional friction”: deliberate moments that help users build confidence before taking action. The idea runs against years of digital design focused on removing steps and accelerating completion, but Camden and Dubois said some AI interactions may produce better outcomes when they slow enough for judgment to catch up.
According to the EY executives, organizations that ignore the tempo gap risk deploying AI experiences that function operationally while failing to earn customer confidence. Their warning is that machine processing can happen instantly, while trust still depends on context, comprehension and human judgment.
This story draws on original reporting from Fortune.