When UX/UI Must Adapt in an Age Full of AI

Type a single sentence, “design a dashboard for an accounting system,” and today’s AI tools will produce a layout, pick a colour palette, set the typography and even write the front-end code in seconds. Work that once took days now takes minutes. The obvious question follows: do we still need UX/UI designers?
Our answer is yes, and perhaps more than ever. But the role is changing fast. When drawing screens is no longer the hard part, the real challenge moves to deciding what to build, who it is for, and why people should trust it. This article looks at five shifts reshaping UX/UI work, and what design teams and businesses should do to prepare.
1. From “screen maker” to “direction setter”
Designers used to spend most of their time on wireframes, colour exploration, spacing and endless mockup versions. AI now does much of this far faster. What it still does poorly is understand business context, organisational constraints, and how real users feel in real situations.
Designers therefore need to move from producing artefacts to directing them: framing the right problem, curating the options AI generates, and choosing the one that best serves both the business and its users. The key skill of this era is not tool speed. It is design judgment, and the ability to explain the reasoning behind every decision.
2. Interfaces are no longer just buttons
More and more digital products now include a chat window, a smart assistant, or an AI agent that acts on the user’s behalf. That changes the design problem completely. Users no longer click through a menu; they state what they want and wait for the result.
Designers must answer new questions. How does a user discover what the AI can do? How easily can they correct a wrong answer? When should the AI act on its own, and when should it ask first? Design expands from screens into conversation, system tone of voice, and invisible sequences of work, a practice often called conversation design or agentic UX.
3. Design for trust, not only for beauty
AI is not always right. It can answer with full confidence while being completely wrong. Part of UX’s job is helping users trust at the right level: not so much that they stop checking, and not so little that they never use it.
Practical patterns include:
- Show where an answer came from, so users can verify it
- Make it clear what the AI is doing right now and how far along it is
- Ask for confirmation before irreversible actions, such as transferring money or deleting data
- Make it easy to undo, edit, or hand off to a human agent
- Explain plainly what personal data the system uses
Products people trust get used again. Products that make people feel out of control get abandoned quickly, however beautiful they are.
4. Personalised experiences that change all the time
AI lets an interface adapt its content, menu order, and even layout to each individual user. Designers no longer design a single fixed screen; they design the rules the system uses to assemble screens.
This is why a design system matters more than ever. Components need a clear structure, with names and usage rules that both people and AI can understand. When AI generates a new screen, the result should still look like the same brand, behave consistently, and meet accessibility standards such as sufficient colour contrast and screen-reader support.
5. Faster user research, but it must go deeper
AI can summarise interviews, cluster feedback, and build test prototypes in a fraction of the time, so teams can test many more ideas than before. What AI cannot replace is talking to real users, watching real behaviour, and understanding cultural context.
Thai users, for example, often have communication styles, expectations around politeness, and mobile habits that differ from the data most models were trained on. Designers who understand users in their local context keep an advantage AI cannot copy.
Skills UX/UI designers should build in the AI era
- Problem framing and business thinking — understand the metrics the business cares about and connect design work to measurable outcomes
- AI literacy — you don’t need to build models, but you should know what AI does well, where it is weak, and how it typically fails
- Prompting and AI-assisted workflows — use AI to explore options, draft UX copy, and build prototypes that actually work
- Close collaboration with developers — as the line between design and code blurs, designers who understand code and design tokens hand off faster and more accurately
- Design ethics — avoid dark patterns, protect privacy and guard against data bias, and respect regulations such as Thailand’s PDPA
Where businesses should start
For organisations planning to bring AI into their digital products, we recommend starting from the user’s problem, not the technology. The first question should be “where are users getting stuck?” before “what can AI do for us?”
Then get the foundations ready: a structured design system, clean and reliable data, and a habit of testing with users continuously. Start with a small feature you can measure, learn from real behaviour, and scale from there.
Conclusion
AI is not replacing UX/UI designers. It is replacing the repetitive work that used to eat their time. That time should go to what humans do better: understanding people, making reasoned decisions, and building experiences users can trust.
At BONMEK, AI is part of how we design and build, but users stay at the centre of every decision. If you are planning to build or improve a digital product for the AI era, talk to our team.


