Gartner recently published a forecast that brings a much needed dose of realism into the customer service AI discussion.

By 2028, more than 50% of customer service organizations will double their technology spend. But Gartner’s key point is not only the increase in technology investment. The more important point is that this increase will not be offset by an equivalent reduction in talent. Gartner also warns that companies trying to reduce headcount too quickly may face operational disruption, degraded customer experience and expensive rollbacks.

That is the uncomfortable part.

Many companies still look at AI in customer service through a very simple equation:

More automation.
Fewer people.
Lower cost.

In practice, that equation will rarely work as cleanly as expected.

AI will change work, not simply remove it

Of course, AI will change customer service.

Simple requests will be resolved faster.
Agents will receive better support.
Knowledge search will become more efficient.
Routing will become more intelligent.
Quality monitoring will become more data driven.

But this does not automatically mean that the human layer in service disappears.

Gartner describes exactly this shift: frontline roles are changing, but they are not disappearing. In its October 2025 survey of 321 customer service and support leaders, only 20% of organizations reported reduced agent headcount due to AI. At the same time, nearly 80% plan to shift at least some agents into new roles, and 84% plan to add new skills to frontline positions.

This is the real message.

AI does not only reduce tasks.
AI shifts responsibility.

The new question in customer service

The most important question for companies is no longer:

“How many people can we replace with AI?”

The better question is:

“What kind of people will we need even more in an AI enabled service model?”

Because when simple cases are automated, what remains is not automatically less important work.

Often, what remains is the more difficult work.

Complex customer issues.
Emotional escalations.
Exceptions.
Quality control.
Knowledge management.
Process correction.
Feedback from real customer conversations.

This is where the new value of human work in customer service appears.

Not in pure ticket handling.
But in escalation, quality, context, judgment and operational control.

Why this matters for DACH companies

Many companies in the DACH region are under three types of pressure at the same time.

Costs need to be controlled.
Customers expect faster and better answers.
Internal teams are difficult to scale.

AI can solve part of this problem. But not all of it.

When companies introduce AI only as a tool, without redesigning the service model behind it, the result is often not a more stable operation. It is a new layer of operational complexity.

Someone still has to check whether AI generated answers are correct.
Someone has to identify where the knowledge base is outdated.
Someone has to decide when a case should not be automated.
Someone has to bring customer signals back into the organization.
Someone has to handle escalations when standard answers are not enough.

These are not low skill service tasks.

These are roles at the intersection of people, technology and operational responsibility.

This is where Southeast Europe becomes relevant

Southeast Europe is often described too narrowly in the outsourcing discussion.

Either as a low cost location.
Or as a talent pool for German speaking agents.

Both views are too limited.

If AI changes customer service roles, then nearshoring also has to be understood differently.

The question is no longer only where companies can find people to process tickets.

The question is where companies can build teams that are able to work in a new service model:

language capable,
fast learning,
technology open,
process close,
operationally well led.

For DACH companies, Southeast Europe becomes relevant when the region is not treated as a cheap fallback option, but as part of a structured operating model.

With clear roles.
With quality assurance.
With knowledge management.
With team leadership.
With escalation logic.
With measurable operational control.

That is the difference between basic outsourcing and a resilient nearshore service model.

The mistake is to put AI against people

The wrong discussion is:

AI or people.

The better discussion is:

Which tasks should AI take over.
Which tasks should remain human.
Which roles will be created or upgraded.
And where can this model be built in a way that is economically, linguistically and operationally sustainable.

Companies that treat AI only as a cost reduction program may be disappointed.

Companies that understand AI as a change in the service model will plan differently.

With less routine work.
With stronger human responsibility.
With better tools.
And with locations that can support this new mix.

Final thought

AI will change customer service. Significantly.

But it will not automatically make customer service people free.

The human layer may become smaller in some areas, but more important in others.

For companies in the DACH region, the question is not only how much technology they buy.

The question is how they organize people, technology and location strategy into one stable service model.

That is where the real decision will be made.

Whether AI in customer service becomes just another cost project.

Or whether it becomes the foundation for a better operating model.