March 05, 2026
Consumers Are Less Forgiving of AI Mistakes Than Human Ones, Medallia Report Reveals
Customer experience teams are deploying artificial intelligence at a pace that would have seemed ambitious just two years ago. Most have measurable goals for it, and many are embedding it directly into customer-facing interactions. What the latest data suggests, however, is that consumer tolerance for what can go wrong has not kept up with the speed of deployment.
Medallia’s new 2026 State of Customer Experience Report, which surveyed more than 500 CX practitioners and 1,500 consumers alongside benchmarks from over 600 anonymised enterprise programs, documents an industry accelerating into AI adoption while struggling with a more fundamental problem, and that is the distance between collecting customer insight and doing something with it.
Consumers Remember When AI Gets It Wrong
More use of artificial intelligence was the single most-cited theme defining CX strategy in 2025, and the vast majority of practitioners currently have measurable AI goals embedded in their planning. Consumers are not opposed to the technology in principle; they accept it readily for routine, low-stakes interactions like checking an order status or asking a simple product question.
The acceptance drops sharply when the interaction carries weight. When disputing a charge, making a formal complaint, or navigating a complex technical problem, the majority want a human, even if it means waiting. On the contrary, when AI handles an interaction and gets it wrong, the consequences are steeper. More than 40% of consumers say human errors are more forgivable than AI errors, while only 7% say the opposite.
Data privacy and response accuracy top the concern lists for both consumers and practitioners, making them among the few points of genuine alignment in the research. According to the report, the strongest use case for AI in CX is not handling customers directly but reducing the friction that makes escalation necessary in the first place, like catching problems early, giving frontline staff better context, and intervening before situations deteriorate.
Loyalty Is More Fragile Than Brands Realise
The AI trust problem is compounding an already serious loyalty crisis, in which 40% of consumers say they have switched a brand they had used for years at least once in the past three months alone. More than half say companies assume they are loyal when they are not. The share describing themselves as very loyal to the brand they most recently transacted with has declined year-over-year.
Brands, for their part, remain largely confident, as most practitioners believe their organisation delivers experiences above the industry average. Only 17% of consumers agree that experiences have improved over the past year. Trust and affordability are among the top areas consumers say are getting worse, not better, a set of concerns that practitioners consistently underestimate.
The Measurement Problem
A significant part of why this confidence persists comes down to what CX teams are measuring and how. Customer feedback surveys and NPS programs remain the dominant data source across the industry, even as response rates continue to fall. Consumers are growing more selective about when they bother: more than half say feedback requests have increased, and a notable share feel companies ask for input without visibly acting on it.
Conversational data from customer service interactions, which is captured in real time where and why experiences break down, is chronically underused despite being one of the most actionable signals available. Seventy-five percent of practitioners acknowledge surveys are insufficient on their own, yet few have moved decisively to expand beyond them.
Insight Without Action Is Just Data Hoarding
The report reveals that a substantial share of departments that receive CX recommendations take no action on them. The problem compounds when CX teams report into leaders outside the Chief Customer Officer or Chief Experience Officer structure, executive follow-through drops considerably in those arrangements, and nearly half of practitioners say their organisation lacks a clear decision-maker with ownership of CX priorities and budget.
Success, meanwhile, is still mainly defined by score movement rather than revenue impact, cost reduction, or churn. Teams that cannot connect experience improvements to financial outcomes are losing ground with leadership, with demonstrating ROI becoming the top-ranked CX priority for 2026, jumping from fifth the previous year
The report closes with three directives for teams that want to stay relevant: expand signals and act on them earlier; move from metric tracking to delivering measurable business outcomes; and advance AI in ways that support rather than sideline the human judgment consumers still demand when it counts.
