Blog 12 · AI in UX design

AI made design output cheap. That is why good judgment just got expensive.

Pranjal Chawan
Pranjal ChawanUX leader · LinkedIn

A designer can now produce ten directions before lunch.

That sounds like a solution until someone asks which one is right.

The argument about whether AI will replace designers misses what is actually happening. The cost of making something is falling. The cost of making the wrong thing is not. A team that can ship faster can also ship the wrong idea faster.

What the 2026 numbers suggest

The UX Tools State of Prototyping: Spring 2026 survey gathered 1,478 responses from designers and builders. It was an open survey shared through newsletters and social channels, so read it as a signal, not a census.

Five of the ten most-used weekly tools were AI tools. Yet only 1.4% said they trust AI output fully without oversight. The largest group, 34.2%, treat it as a first draft they edit heavily.

The top blockers are telling as well. Time to learn tools came first at 55.7%, then too many tools at 53%, then output quality at 52.2%. Adoption is the easy headline. Making it work well is the harder job, and it is a leadership job.

Where I use AI, and where I do not

StageAI helps withA human must own
ResearchSummarising, clustering, drafting guidesWho we ask, what it means, what we believe
StrategyExploring options and counterargumentsThe bet, the trade-offs, the accountability
DesignVariations and quick prototypesFit, taste, accessibility, emotional tone
DeliveryDocumentation and repetitive tasksQuality standards and final review

My rule is simple. AI can speed up the work, but it cannot be accountable for the decision.

On an enterprise modernisation project, we combined AI-driven workflows with reusable components and WCAG-compliant patterns, and application load times fell by 20%. AI was one part of that system. The standards around it did much of the heavy lifting.

Three risks UX leaders should own

Confident nonsense. AI summaries can sound certain while missing the one comment that mattered. Go back to raw data for important decisions.

Beautiful sameness. If everyone uses the same tools and prompts, products begin to look alike. Your insight into your own users is what keeps you different.

Skipped thinking. Fast output can tempt teams to jump over research. Remember my Fixed Deposit project: we built a thorough journey for the wrong audience. A faster tool would have helped us reach that mistake sooner.

Designing products with AI inside

When AI is part of the product itself, the design questions change:

These are trust questions before they are technical ones. Good AI experiences set clear expectations, leave people in control, and make recovery easy.

My advice to design teams

Give people time to learn, not another tool. Agree on where AI can and cannot be used. Share what worked. Keep research close to real people. Treat your design system as the structure that keeps AI output consistent.

I keep learning here too. Recently, I completed Generative AI for Everyone, an AI tools workshop and the Generalist phase of House of Edtech's AI programme, and I remain very curious about where this is going.

Questions people ask

Will AI replace UX designers? AI will change what designers spend time on, but people still need to set direction, judge quality and answer for decisions.

How can UX designers use AI responsibly? Use it to speed up low-risk work, review outputs carefully, protect user data, and keep humans accountable for decisions.

How is AI changing your team?

I would love to compare notes, especially on what has not worked. Write to me at pranjal.productmind@gmail.com or connect on LinkedIn.

© 2026 Pranjal Chawan. All rights reserved.