Customer service AI in 2026: more automation, more people

Executive summary

The pressure to deploy AI in service is close to universal. 91% of customer service leaders in a Gartner survey of 321 leaders said executive leadership is pressing them to implement AI.[1] AI rose from tenth to second place among service leaders’ priorities in Salesforce’s latest survey.[6]
Money is following the pressure. Service leaders’ spending on AI rose 38% while their overall budgets grew 2%, according to a Gartner survey of 199 leaders in April and May 2026.[3]
Automation is real but partial. Service teams estimate that AI handles 30% of cases today and expect 50% by 2027.[6]
The mass layoffs have not arrived. 85% of service leaders are adding responsibilities to human agents, and 31% have made or plan frontline reductions through the first quarter of 2027.[2] Gartner expects half of the organisations that planned big AI-driven cuts to drop those plans by 2027.[4]
Customers are wary. 50% of consumers worry AI will keep them from reaching a human, and 53% fear misuse of their personal data when companies automate interactions.[7] 54% of US customers trust human agents more than AI for recommendations, against 32% who trust AI more.[2]

Key findings

0191% of customer service leaders report pressure from executive leadership to implement AI

02AI spending in service rose 38% while overall service budgets grew 2%

03Service teams estimate AI handles 30% of cases today and expect 50% by 2027

Customer service cases handled by AI

Today

30%

Expected by 2027

50%

Source: Salesforce, November 2025

Who US customers trust more for recommendations

Human agents

54%

32%

Source: Gartner, 2026

What leaders should do

Redesign the highest-volume journeys end to end before automating them.
Measure resolution and effort, not handle time alone.
Invest in knowledge management as the foundation for automation.

In the full report

Pressure from the top, money from the AI line
How much work AI is doing
People: more responsibility, fewer layoffs than forecast
Customers: willing to use AI, unwilling to be trapped by it
What separates useful automation from deflection
What to do in 2027
Sources

Methodology

ConfX research brings together our own analysis, what we hear from senior leaders at ConfX summits, executive roundtables and briefings, and the latest published research from regulators, academic institutions, industry bodies and research firms. Analysis completed on 15 October 2026.

Every figure is attributed to its source. Samples, regions and definitions differ between studies, so figures from different sources describe different groups and should not be combined.