What the 2026 evidence says about AI in customer service

Pressure and money arrive together

Gartner surveyed 321 customer service leaders for its 2026 priorities research: 91% said they were under pressure from executives to implement AI this year, and the top priorities they named were customer satisfaction, operational efficiency and self-service success.¹

A second Gartner survey of 199 leaders, published in August, found that AI spending in service and support rose 38% while the function’s total budget grew 2%.² The money is coming out of existing lines. Leaders expect GenAI chatbots, voicebots and agentic AI platforms to deliver the most value over the next two years.²

Human agents stay in the model

The expectation of mass layoffs has not shown up in the data. In April 2026 Gartner reported that 85% of service and support leaders are expanding the responsibilities of their human agents, and 31% have implemented or plan AI-driven frontline layoffs through the first quarter of 2027.³ Gartner describes this as workforce redesign.

Its earlier forecast points the same way: no Fortune 500 company will have fully eliminated human customer service by 2028.⁴ The routine work moves to AI, and the harder cases stay with people who now need broader skills.

What customers ask for

Two 2026 surveys describe the customer side. 87% of customers say that companies using GenAI for service must provide access to a human agent.⁵ And only 27% would try a chatbot again after a negative experience.⁶ A failed automated conversation costs more than one contact: it changes how the customer uses the channel afterwards.

Customers are arriving with their own assistants

A July 2026 Gartner survey of 3,566 customers found they are three times more likely to use third-party GenAI tools than company-provided chatbots for service, and 58% of customers who use GenAI have had it complete a task for them.⁷

This matches a forecast Gartner made in December 2024: by 2028, 70% of customer service journeys will begin and be resolved in conversational, third-party assistants built into mobile devices.⁸ For a service organisation, the first contact may increasingly come from software acting for the customer. Our report on agentic commerce looks at the payment side of the same shift.

The cost line

Gartner’s best-known forecast is that agentic AI will autonomously resolve 80% of common customer service issues without human intervention by 2029.⁹ Its January 2026 forecast adds a cost warning: by 2030, GenAI cost per resolution will exceed $3, more than many offshore human agents. Gartner cites rising data centre costs, vendors moving from subsidised growth to profitability, and more complex use cases.¹⁰

Read together, the forecasts describe a service model in which AI resolves more and costs more per resolution than vendors’ per-unit prices suggest. The business case has to count the whole operation.

Questions for the 2027 plan

1. Which cost line funds the AI budget? Name it before the spend is approved.
2. What is a resolution? Agree a definition that includes the customer not having to come back.
3. Where does the customer reach a person? Map the route from every automated channel.
4. What happens when the customer is an agent? Decide how you authenticate and serve a third-party assistant.
5. What does each resolution cost in full? Include model usage, integration, quality review and escalations.