Five reasons design teams are struggling to adopt AI
Reinventing your methods, process and tooling is easier said than done.
AI is turning the design process upside down, yet a lot of design teams have barely gotten started on their journey to rethink how they work.
Although design tooling is still immature, it’s clear that the way that designers will be working in 5 years’ time is completely different to how they work now.
The technology is here, but many teams are moving slowly to make the wholesale changes that are required. The gap between what is possible and what most teams are actually doing is huge. Why?
1. No-one on the team is technical enough
Before AI, you didn’t hire designers for technical skills – it was a nice to have. The problem is that some teams don’t have anyone who is technical.
No-one who is tinkering with Claude Code in their evenings. No-one who understands the terminal, git or how the thing they’re designing actually gets built. No-one to lead their thinking around changes to tools, ways-of-working and methods.
In the UX Tools State of Prototyping survey, the designers doing most of their building with AI were overwhelmingly the ones who already had an engineering-adjacent background. The people best placed to build with AI are the ones who already have a technical background.
You don’t need designers to code like engineers, but you do need enough technical grounding to make sensible calls about tools, workflow and how design joins up with development. If nobody has that grounding, where does the new process come from?
2. You have to bring developers with you
New ways of working mean that designers and engineers have to work closely together. The ideal is that you’re in the same code repository as the engineers, not throwing a Figma file over the wall.
For most teams that’s a long way from today. Designers and developers might sit in the same squad, but they’re still siloed, with a big handover in the middle.
So when you’re trying to work out what your AI design process is going to look like, you’re not just choosing new tools for yourself. Your engineering partners have to adopt the same approach, because it makes no sense for you to work one way and them another.
If you’re starting from a position where design and engineering are already not aligned, and are used to working in their silos, then there’s a lot more to do than just choosing a tool.
3. AI’s usefulness to design is uneven
In the old way of working, the output was roughly the same whatever the project: a high-fidelity Figma file.
Now it varies wildly. If you’re assembling a feature on an established product with a mature design system, you barely need Figma at all. AI is good at using existing components and you’re not asking it to do any visual design, which it’s bad at. You can sketch the rough idea, have it assemble it and iterate. There’s little point building a pixel-perfect version by hand first.
But if you’re working on a new product which needs its own visual design or brand, it’s a different job. AI can get to a rough wireframe using generic components, then it stalls. Creative and original visual design still needs a human.
So the impact of AI on the design process is uneven. It transforms one type of project and barely touches another. That makes scoping harder: you can’t assume every project takes the same shape, needs the same skills or moves at the same pace any more.
4. AI isn’t good at design!
I could write a whole newsletter on this, but the short version is... AI is nowhere near as good at design as it is at coding. With code there’s a verifiable right answer - does the test pass? With how an onboarding flow should work, there isn’t a ‘correct’ answer. Design requires human creativity and judgment.
If you ask AI to design something, you tend to get a result that looks right but has no thinking behind it. It copies the surface and skips the reasoning. Then you have the problem that it’s terrible at novel visual design.
At least with the latest models, AI creates functional but generic design work. You get the average of what it’s seen. Having it use an established design system to assemble pages for you gets better results because you’re not asking it to be creative or exercise any aesthetic judgement.
As with any task, you get better results from AI when you give it a more limited task with a narrower scope. If you break the work into smaller steps and you do the decision-making, then it’s ok. But this also means that adopting AI into your design process isn’t a plug-and-play task.
5. Nobody has time to work it out
Figuring out all of this takes time, and most teams don’t have any spare.
Four years ago, picking a design tool took about five minutes: it was Figma, like everyone else. Now there are countless options and they change every month. There’s no settled answer to copy.
Working out which tools to use, how to align with engineering and what your new process looks like is a lot of work. It requires an investment of time and resources to do, which many stretched design teams can’t afford.
This can‘t be done off the side of the desk
Most teams need help to get unstuck: new hires with technical skills, outside expertise or leadership carving out protected time. You can’t figure this out properly while delivering everything else at the same time.
The teams adapting fastest aren’t more aware of the need to change. They just have a head start: technical people already on the team and close existing relationships with engineers. If you’ve got neither, it’ll take longer.
The best way to figure this stuff out is to build something together as a proof-of-concept. Only through learning by doing can you figure out what’s going to work for your team. But doing so needs dedicated time.
