Everyone has been trapped in a chatbot loop. You type "agent," it offers you a help article. You type "AGENT," it asks if you meant billing. For many businesses, AI at the front desk or on the support line looks like an easy cost cut. The 2026 research says customers will accept it, but only on one condition, and businesses that ignore that condition are paying for it in trust.
What happened
In August, Gartner released a survey of 3,566 business and consumer customers, taken in February and March 2026. As CMSWire's reprint of the release lays out, half of customers said AI makes their service interactions easier. But 87% said any company using generative AI for customer service must give them a way to reach a human agent. One more finding deserves attention: in their latest service interaction, customers were about three times more likely to use an outside AI tool like ChatGPT or Gemini than the company's own chatbot. Gartner analyst Eric Keller said companies "should not use GenAI as a mandatory first step for every issue."
Meanwhile, the people running service teams are under heavy pressure to automate. Earlier in the year, Gartner reported that 91% of the customer service leaders it surveyed felt pressure from senior executives to implement AI in 2026, and that nearly 80% of organizations planned to move at least some agents into new roles.
On the customer side, a May survey released by AnswerConnect, a company that sells human-staffed answering services, polled 6,000 consumers in the US, UK and Canada. It found that 85% preferred to speak with a real person, up from 83% six months earlier. Frustration with AI agents rose from 54% to 59%, and 57% said their trust in a business would drop if it relied mainly on AI for service. AnswerConnect has an obvious interest in that result, so weigh it accordingly. Still, it points the same way as Gartner's independent data.
Why it matters for your business
Customers aren't against AI. They're against being trapped. Half of Gartner's respondents said AI makes things easier. The anger comes from bots that block the way out, not from bots that answer simple questions well.
Your customers bring their own AI now. If people are three times more likely to ask ChatGPT about your product than to use your chatbot, then what those tools say about you matters. Clear, accurate information on your website feeds those answers.
For a small business, every call is a relationship. A large bank can lose a few customers to a bad bot and barely notice. A dental office, law firm or contractor can't. Your phone line and front desk are part of what people are buying.
After-hours is where AI earns its keep. Missed calls at 8 p.m. are lost business. An AI that answers, books appointments and takes messages when no one is there adds something you didn't have before, instead of replacing someone you did.
To be fair to the push for automation: wait times and staffing costs are real, and plenty of customer questions really are simple. Store hours, order status and appointment changes don't need a person. The mistake is treating every contact as if it were one of those.
What we'd do
- Make the human exit obvious. One clear phrase or button gets a person, or a guaranteed callback. Never hide it.
- Start with after-hours and overflow. Use AI where calls are going unanswered now, not where a good employee is already doing the job.
- Limit AI to what it does reliably. Hours, directions, bookings, order status, intake questions. Hand off anything emotional, complicated or involving money.
- Pass the context along. When AI hands a customer to a person, the person should see what was already said. Making people repeat themselves undoes the benefit.
- Tell people it's AI. Disclosure builds trust. Pretending a bot is a person destroys it when they find out.
- Fix your public information. Update your website, hours and FAQs so outside AI tools give customers correct answers about you.
- Read the transcripts monthly. Look for where people got stuck or asked for a human, and fix those paths.
Our view: AI at the front desk is worth doing, but as a helper, not a wall. The research is consistent that customers will happily let a machine book their appointment, as long as a person is one step away when it matters. For a small business, that human option isn't a cost to minimize. It's the part of your service people remember.
Want a system like this in your business?
A discovery conversation is where we map it: what to automate first, where the human checkpoints go, and what it would take.
Book a discovery conversation →