“AI is coming for every design job” makes a dramatic headline. It is a less impressive piece of workforce analysis.

The evidence supports a more careful question: which tasks change, and what happens to the people learning those tasks?

Jobs and Skills Australia’s national Gen AI Capacity Study concluded that generative AI is more likely to augment human work than replace it outright. Some routine tasks may be automated, while the study expects broader changes in how work is performed and what skills are valuable.

That distinction matters for UI/UX.

Design work contains a lot of tasks AI can accelerate: moodboards, content variants, layout exploration, research synthesis, accessibility checks, prototype generation and even production code.

But UI/UX is not only those tasks.

The role also involves understanding ambiguous requirements, resolving competing stakeholder interests, deciding what not to build, observing real users, negotiating technical constraints, communicating trade-offs and taking responsibility for the resulting experience.

Those responsibilities are much harder to automate cleanly.

The available labour data gives context, but it cannot settle the future of UI/UX jobs.

Jobs and Skills Australia’s older ANZSCO profile reports 203,200 Software and Applications Programmers and annual employment growth of 13,700. Those figures describe programmers, not UI/UX designers; the profile is no longer updated. They cannot be used to forecast design employment or prove that AI will preserve every digital role.

Graphic and web design occupations face a different pressure profile, but even there the question is less “does design vanish?” and more “which parts of design remain economically valuable?”

The answer is likely to move upward in abstraction.

Producing a standard interface becomes easier.

Understanding the system becomes more valuable.

Designers who can combine visual quality with research, product strategy, accessibility, design systems and technical fluency should be better positioned than people whose role is limited to assembling predictable screens.

The same is true in development.

AI can produce code quickly. It does not remove the need to understand architecture, security, maintainability, browser behaviour, infrastructure or the consequences of a change.

Some jobs may change substantially, and entry-level production work may be compressed. The scale and distribution of those effects remain uncertain.

But “AI replaces designers” is too crude.

A more useful framing is:

AI can automate parts of jobs. Employers then decide whether the saved time expands human work, changes the skill mix, or reduces hiring.

That makes adaptation the real employment issue.