The apparent customer backlash against AI in support is a measurement problem masquerading as a technology problem. Recent research showing declining preference for AI-driven support obscures a more nuanced reality: customers accept automation readily when it solves their problem and resent it when it creates friction. The AnswerConnect data revealing four in five customers find AI helpful for simple questions, coupled with Glance's 2026 CX Trends finding that 68 percent prioritise complete resolution over speed, indicates the issue is not the technology itself but how organisations deploy it. The real culprit is operational design—specifically, the widespread practice of optimising for deflection rates rather than resolution rates. This distinction matters enormously for teams already running AI-native platforms like Agentforce or Freshdesk's automation layers: a bot that contains a contact by making escape difficult, or that closes a session without solving the underlying problem, generates the same customer contempt as a poorly trained human agent. The backlash reflects decades-old service principles applied to new tools. Customers object to repeating themselves across channels, hitting walls with no visible exit to a person, and facing associates who demand they restart their explanation after an automated handoff. None of these grievances are about artificial intelligence; they are about friction, context loss, and the perception of being trapped in a system designed for operational convenience rather than customer outcome.
The operational seams where automation meets human support determine whether AI builds or destroys trust. The most expensive mistake leaders make is rewarding containment metrics that hide poor resolution rates behind green dashboards, only to watch customers defect quietly after calling back twice. When organisations shift measurement to resolution and customer effort, automation improves because the contacts reaching associates are genuinely complex cases where judgment and empathy create value. The handoff is where most damage occurs—a customer reaching a live associate only to be told to start over erases the value of the preceding interaction and signals that the organisation's systems do not communicate. Fixing this single operational seam, ensuring the associate opens with full context of what the customer attempted and why it failed, outperforms most automation investments. For support leaders evaluating whether to expand AI deployment, the question is not whether to automate more or less, but whether your current infrastructure passes context reliably between channels and between machine and human. Financial services environments, where tolerance for loops is near zero and emotional stakes are high, reveal the distinction clearly: customers welcome invisible AI—routing algorithms, fraud detection, account history surfacing—that informs and accelerates human judgment, whilst they resent visible AI that blocks access to resolution. This principle holds across industries. The winners over the next several years will not be those using the least or most automation, but those using it to make service feel effortless and human, in that order. The $1.9 trillion in US consumer spending at risk from poor experiences, with four in five customers switching brands after a bad one, remains unchanged regardless of whether the bad experience came from a bot or a person.
Why Customers Don’t Hate AI. They Hate Bad Service. CustomerThink