Gartner says that by 2028, more than half of customer service organizations will double their technology spending, and that increase will not be matched by an equal drop in staffing. The firm’s latest forecast points to a pattern: companies are adding AI and automation tools, but they still need people to run, supervise and fix those systems.
According to Gartner, organizations are under pressure to use AI to reduce labor expense, but rapid headcount cuts can create operational disruption, weaken customer experience and force expensive reversals. In other words, the technology bill is rising faster than the savings.
“Leaders are hoping that AI will deliver immediate cost savings, but most organizations are understating the talent required to make AI successful,” said Kathy Ross, Vice President Analyst in the Gartner Customer Service & Support practice. “Technology spend is rising rapidly, yet talent needs are evolving – not disappearing.”
The clearest takeaway from Gartner’s prediction is that AI is not replacing the customer service workforce at the pace many executives expected. Gartner’s survey of 321 customer service and support leaders, conducted in October 2025, found that only 20% of organizations had reduced agent headcount because of AI. That suggests most companies are still in the early phase of adoption, where technology is being layered on top of existing service operations rather than replacing them.
Gartner says nearly 80% of organizations plan to shift at least some agents into new roles, and 84% plan to add new skills to frontline positions. Those changes require training, workflow redesign, integration work and support systems. AI may lower some manual tasks, but it also creates a new workload in administration, governance and quality control.
The report also shows that organizations are still using staffing changes to fund technology, not to eliminate labour altogether. Gartner’s analysts said many leaders are cutting agents to pay for AI, rather than because AI is fully mature enough to take over the function. That is a significant distinction for enterprise planning. It means customer service spending is shifting from payroll-heavy budgets to software-heavy budgets, but the overall cost base is not falling in the near term.
Why customer service costs are rising
Customer service costs are rising because AI is adding layers, not removing them. The technology stack now includes conversational AI, routing tools, knowledge systems, analytics, orchestration and oversight mechanisms. Each of those tools carries licensing, implementation, maintenance and integration costs. Gartner’s forecast suggests that this mix will continue to expand through 2028 rather than compress quickly.
There is also a practical reason headcount is not dropping fast. Customer service work is messy. Many issues still require judgment, exception handling and escalation. Gartner’s warning that rapid reductions can lead to degraded customer experience and rollbacks is a useful reminder that service automation has limits. Companies can automate routine interactions, but they still need humans for complex cases, exception handling and recovery work.
What this means for customer service leaders
For service leaders, the message is not to avoid AI; it is to budget for a longer transition than many board decks assume. Gartner’s forecast implies that AI in customer service is becoming a capital and operating expense simultaneously: capital-like in implementation and software investment, operating-like in training, governance and ongoing oversight. That makes cost control more difficult, not less.
The report also suggests that the biggest near-term gains will come from workforce redesign, not workforce elimination. Organizations that use AI to support agents, improve routing and reduce repetitive work are more likely to get measurable returns than those trying to remove large numbers of people too quickly. Gartner’s data on role shifts and skill additions points in that direction.
Gartner’s forecast is less about a tech spending spike than a structural change in customer service economics. AI is changing how the function works, but it is not yet making the function cheaper in a simple, linear way. By 2028, many companies will spend far more on customer service technology, while still carrying much of the human cost that existed before. That is why customer service budgets are rising, and why the shift is likely to stay expensive for some time.
