Consumer Attitudes Toward AI Customer Support
What 1,083 consumers across the US and Europe really think.
Most consumers say they prefer human support, but their actual choices say something different.
In our survey, 68% of consumers say they prefer a human. However, when we introduce a real tradeoff, that preference doesn't hold up. This boils down to three core findings.
6×
Consumers are 6× more likely to prefer AI when there's a wait
Baseline preference for AI is 7%. When the choice is "wait 15 minutes for a human or get help from AI right now," 44% switch to AI. 24 points move directly from team-human to team-AI the moment time becomes a factor.
3×
Consumers are 3× more likely to prefer AI for embarrassing or personal topics
When the topic is sensitive — money, health, security — 21% prefer AI and another 40% don't care which channel they use. 61% will share sensitive information with AI over a human.
52%
Good AI experiences are brand-accretive
52% of consumers say they've had a genuinely good AI support experience. Of those, 50% now think more highly of the brand (only 3% think less). Every successful interaction is brand equity.
In this report we present the supporting data and what these findings mean for CX leaders in fintech, insurance, and healthcare.
Why we did this
We're interested in doing right by consumers. Consumer preferences in complex industries like fintech, insurance, and healthcare can look different from preferences in simpler categories, and most published research is on the simpler ones.
Went to the source and surveyed 1,083 consumers across the US and Europe to triangulate what we already see working with leading brands like Airwallex, Arbor, Breeze, EasyKind, Eucalyptus, Flex, LinkTree, and Talkspace.
The model most companies are running today doesn't work for customers. A chatbot handles the easy questions, then bounces everything else to a human queue. The customer ends up where they started, often after explaining the problem twice. Same wait, same outcome, plus the friction and frustration of being shuffled. The data here points at a sharper question: where in your operation does resolution time, accuracy, and 24/7 availability matter most to the customer?
Methodology
- Sample: 1,083 consumers
- 8 countries: US, UK, Germany, France, Netherlands, Ireland, Spain, Italy
- Survey platform: Prolific (verified panel)
- Data collected: April 7, 2026
Finding 1 — Wait time is the lever
Stated preference vs actual choice
| Question | Prefer human | No preference | Prefer AI |
|---|---|---|---|
| Do you prefer human or AI support? | 68% | 25% | 7% |
| Would you rather wait for a human or get help from AI right now? | 56% | — | 44% |
Stated preference for humans sits at 68%. The moment a wait is on the table, AI preference jumps from 7% to 44%. The 24 points that flip are the most actionable number in this report.
Tolerance for waiting is razor thin
52.5% of consumers abandon a human queue within 15 minutes. 75% within 30. Only 16% will wait as long as it takes. Most enterprise CX queues fail well inside that tolerance window.
Finding 2 — Privacy is an AI advantage
When sharing sensitive information, do you prefer AI or human?
39%
prefer human support
21%
prefer AI support — 3× the 7% baseline
40%
have no preference
61% of consumers either prefer AI or don't care when sharing something sensitive. AI doesn't judge, doesn't gossip, and handles data the same way every time. For CX leaders deciding where to deploy AI in regulated categories, this is the data point that should reset the prior.
A related signal: 95% want agents that remember
| How valuable is an agent that already knows your full history? | % |
|---|---|
| Very or extremely valuable | 70% |
| Somewhat valuable | 25% |
| Don't care | 5% |
AI agents now remember every interaction, while human agents simply can't. In categories where context determines whether the response is right, that gap compounds.
Finding 3 — AI support is brand equity
Have you had a genuinely good AI support experience?
Did a good AI experience change how you view the brand?
The "never tried it" group has shrunk to 4%. Half of those with a good AI experience now think more highly of the brand. Each successful interaction shifts the baseline.
Why people say they prefer humans
The dominant reason people give for preferring a human is "I want someone who understands my situation." That is a need that a context-aware AI agent is built to serve.
86% of these objections are about capability, trust, and context. Categories like these are what current context-aware AI is designed to address. Only 8% are anchored in an actual bad AI experience. The remaining preference for humans is mostly reputational, and reputation moves with each successful interaction (see Finding 3).
As complexity rises, human preference rises
Consumers fine with AI support, by scenario
71%
Password reset
53%
Refund on cancelled order
45%
Insurance coverage check
42%
Medication side effects
38%
Loan / repayment options
29%
Disputing a charge
Even in the most sensitive scenarios — disputes, loans, medication — 29–42% of consumers are still fine with AI. The complexity gradient is real, and AI is never rejected outright on it.
"I don't care if it's AI or human, as long as my problem gets solved."
58% agree
In practice
The data describes attitudes. Here's what happens when CX teams deploy AI in the scenarios consumers say they're most skeptical about.
Insurance
SageSure is an MGU with 850,000+ policyholders in catastrophe-exposed residential property. AI now reads the email, updates the policy, and replies to the agent without a human touching it.
Pete Rizzo
AVP, SageSure
SageSure
Healthcare
Eucalyptus is a digital health company across weight management, men's health, and reproductive care. As volume tripled, AI took repeatable clinical workflows. CSAT improved by 10 points.
3× ticket load
CSAT up 10 points
Eucalyptus
"Not only can Lorikeet receive an email on my behalf, they could update the policy, and they could reply to the customer without me ever needing it to cross the desk of one of my employees."
Pete Rizzo · AVP, Service and New Business Onboarding Optimization, SageSure
Implications for CX leaders
01
The resistance is softer than the headline.
68% "prefer humans" reads like a wall. 32% are already neutral or pro-AI before any tradeoff. Another 24 points flip the moment wait time is on the table.
02
The objection is solvable.
"Want someone who understands" — the top reason consumers give for preferring humans — is a context problem context-aware AI is designed to solve. Only 8% of objections come from an actual bad AI experience.
03
Queue time is the leverage point.
52.5% abandon within 15 minutes. If a human queue runs longer than that, most customers who said they prefer humans will choose AI when offered.
04
Sensitive topics are not off-limits.
61% prefer AI or don't care when sharing sensitive information. The privacy advantage is a real consumer preference.
05
Good experiences compound.
50% of consumers who've had a good AI experience now view the brand more positively. Brand equity rides on what AI actually does, not the channel choice.
The numbers that matter
6×
more likely to prefer AI when there's a wait (7% → 44%)
3×
more likely to prefer AI for embarrassing or personal topics (7% → 21%)
52.5%
abandon a human queue within 15 minutes
58%
agree the outcome matters more than the channel
95%
want agents that remember their history — context is the killer feature, and only AI can deliver it consistently at scale
71%
fine with AI for account security
61%
fine sharing sensitive info with AI