In 2026, 91% of customer service leaders are under strong pressure to implement AI. This is the finding of a recent Gartner analysis.
Many companies are not introducing AI because they are strategically ready.
They are introducing it because they believe they have to. That is a problem, and the pressure is very real.
A market observation: AI is moving from a differentiator to a standard expectation.
This changes expectations at several levels: customers expect automation to work, decision-makers expect measurable efficiency, and markets expect scalable solutions.
Using AI alone therefore says less and less. What matters is how well it is implemented.
The technology is often not the problem
The core problem rarely lies in the technology itself. It lies in the foundations behind it.
In many organisations, AI is added to processes that have never been clearly defined. Responsibilities are unclear, handovers along the customer journey break down, and information is fragmented. AI takes on these structures and reproduces them.
At the same time, the role of data is systematically underestimated. Without a consistent data foundation, companies produce contradictory answers, inconsistent service and a customer experience that cannot scale.
There is also a strategic imbalance: efficiency is prioritised, while customer experience is often only an implicit consideration. This leads to solutions intended to reduce costs without creating real value for customers.
In many cases, governance becomes relevant only when problems arise. A lack of transparency, traceability and attention to regulatory risks is addressed too late—a particularly critical issue in regulated industries.
If processes are already poorly defined or unclear, AI makes that same confusion affect more customers, more quickly.
A system that accesses poor data does not deliver better answers. It delivers incorrect answers faster.
The pressure is real
The Gartner analysis makes this pressure visible. AI in customer service is no longer optional; the direction of development is clear.
However, what matters is the quality of implementation, rather than how quickly it is introduced.
AI does not solve structural problems. It reveals them and can amplify them.
Between ‘We must use AI’ and ‘Our AI measurably improves customer service’ lies a gap that many companies currently still underestimate.
The decisive question is therefore not:
Are we using AI?
The decisive question is:
What exactly are we scaling, and on what data foundation?