Artificial intelligence has long been part of digital products. Nevertheless, in the context of UX, it is still surprisingly often limited to a single application: the chatbot.
The logic behind this is understandable. Chat interfaces are visible, easy to integrate and quickly convey a sense of innovation. In recent years, many companies have introduced chatbots to automate customer service or handle simple support requests.
But this perspective falls short.
AI is not only changing individual points of interaction. It is changing the structure of digital experiences themselves. Interfaces are becoming more dynamic, systems more context-sensitive and user interactions increasingly dialogical. At the same time, new forms of interaction, new user roles and even new target groups for digital products are emerging: AI agents.
UX is thus facing a fundamental expansion of its previous understanding.
Voice interfaces as a new layer of interaction
One of the most notable developments is the resurgence of language as an interface.
Voice assistants, conversational interfaces and multimodal systems make interactions more natural and immediate. Users describe problems or wishes in their own words instead of navigating through menus.
This approach has enormous potential, particularly in complex decision-making processes.
One example is financial products. Insurance, loans and pension plans are often difficult for many users to understand and are associated with uncertainty. Traditional product pages with tables, forms and comparison calculators often create more complexity than clarity.
AI-based advisory interfaces can simplify this complexity. They analyse individual requirements, explain options, simulate scenarios and guide users step by step through the decision-making process.
The interface becomes less of a surface and more of a dialogue-based advisory system.
AI agents as digital actors
This change is becoming even more profound as autonomous AI agents become more widespread.
These systems are no longer act merely as assistants within an interface. They can perform tasks independently: researching information, comparing offers, making bookings or preparing transactions.
This changes the role of the user. Decisions are increasingly being delegated
For example: an AI agent can analyse travel options, compare prices, check reviews and suggest an optimised route. The user simply confirms the final decision.
Digital systems therefore have to function not only for human users, but also for machines acting on their behalf.
AX – Agentic Experience
AX describes the design of digital systems for interacting with AI agents. While UX traditionally considers humans as the primary users, AX extends this understanding to include a second user group: autonomous systems.
For companies, this means new requirements for digital touchpoints.
Information must be both visually understandable and machine-interpretable. Product data, availability, prices and contract terms must be structured and easily accessible.
Websites are thus evolving from interfaces to APIs for intelligent systems.
Agentic Commerce
One of the most visible consequences of this development is agentic commerce.
AI agents are taking over parts of the purchasing process. They research products, compare offers, evaluate alternatives and make preliminary decisions.
This is fundamentally changing the classic e-commerce funnel.
Rather than guiding users through interfaces, brands must ensure that their products are present in the decision-making logic of agents. Product information, reviews, delivery terms and price structures are becoming machine-readable decision-making factors.
In this context, UX is expanding to include a new dimension: the design of decision-making bases for autonomous systems.
Competition is therefore shifting partly from shop interface to the agents’ algorithms.
The future of UX
For a long time, UX was the discipline of interfaces. Navigation, layout, interaction design.
With AI, UX is increasingly becoming the discipline of intelligent interaction.
Design now concerns not only screens, but also dialogue, decision-making logic and automated processes. Users interact not only with interfaces, but with systems that can understand, learn and act.
