Where is the design profession and customer-centred thinking heading in an AI-accelerated world? – Part 1 (artificial intelligence)

When design is finding its place again

Thoughts on design roles, AI and customer-centred ways of working at the beginning of 2026 (artificial intelligence)

Introduction

I have worked in and around customer-centred design for more than fifteen years. I have seen UX and CX evolve from a niche, long-marginal approach into a core principle of digital operations. Research, testing, workshops, journeys – these have all become everyday tools, and we rarely need to explain why the customer perspective matters. (user experience) (customer experience)

And yet, lately I have increasingly felt that we have stalled.

Not because UX does not work. Quite the opposite: the methods are more mature, faster and more accessible than ever. The question is whether the same answers we learnt in a simpler environment are enough for today’s more complex operational challenges. (user experience)

This is not a list of trends. Rather, it is an analytical reflection on how the role of design is changing in a more complex, uncertain environment – and why this has once again become a relevant question.

1. Design searching for its role in a more mature environment

Design is now present almost everywhere. It is part of product development, it appears in organisational language, and its legitimacy rarely needs to be justified. Yet in many organisations, its role remains difficult to define.

If someone works in sales, IT or marketing, it is generally clear what that means: which decisions they influence and what they are responsible for. With design, this is far less clear. The same job title can represent very different roles in different places.

This partly stems from the nature of design: it is a cross-disciplinary profession that brings together business, technology and customer perspectives. The problem arises when this position at the intersection is not a conscious decision but an established practice: design is ‘there’, it is involved, but it is unclear when and on which issues it carries real weight. In such cases, design can easily remain a support function: helping to make the case, make decisions easier to understand and reduce friction – but rarely shaping direction. For a long time, this was not critical. Digital products used to be simpler, appeared across fewer channels, and the impact of decisions often remained within the boundaries of an individual team or function. Today, the digital experience is created across multiple channels through the collaboration of several teams and systems, while customers’ expectations and sensitivity have also increased. What is more, an increasing number of decisions extend beyond a single project or organisational unit. In this environment, a purely supportive role for design is no longer enough. Not because design wants more, but because the environment places greater weight on the questions it has traditionally addressed: connections, consequences and operations understood from the customer’s perspective.

2. AI is not a tool, but an operational catalyst (artificial intelligence)

We often talk about AI as a new tool or feature. In reality, however, it rarely remains within the boundaries of a single team or system. As soon as it appears, it brings questions to the surface that often had not even been considered before: who can use it, what decisions it can make, who is accountable, and how it fits into the overall operation. (artificial intelligence)

These questions affect the business, IT, legal, customer service and customer experience teams at the same time. AI is therefore not a technological issue, but an operational issue. It is a catalyst: it amplifies organisational tensions that were previously less visible. (artificial intelligence)

This is where the service design perspective becomes especially relevant. Not as a methodology, but as a framework for thinking: one that connects different perspectives and makes the consequences of decisions explicit.

3. Experimentation as a responsible implementation model

An AI implementation is expensive, uncertain and difficult to reverse – both technologically and organisationally. The ‘big bang’ approach is therefore particularly risky. Instead, experimentation, and particularly the POC (Proof of Concept) as a learning methodology, is increasingly coming to the fore. (artificial intelligence)

Here, a POC is not a technology demo, but the validation of a hypothesis: whether AI provides a genuinely meaningful answer to a specific operational problem. Validation is not only technical, but also organisational, legal and customer-focused. The aim is not rapid implementation, but learning and building buy-in. (proof of concept) (artificial intelligence)

This approach clearly shows that design – particularly in the service design sense – does not accelerate; it provides a framework. It helps ensure that experimentation is not an ad hoc attempt, but a deliberate path to implementation.

Part two: When execution no longer sets you apart

The author is Ákos Maczinkó, Managing Director of Works.