A Principled Take On the Use of AI in UX

My philosophy

I view AI as a way to supercharge my creative process, not replace it. For me, AI is a strategic accelerator. It helps me ease the friction caused by tasks that aren’t my core strength and passion, freeing me to focus my energies on high-value creative and strategic work.

I use it to quickly analyze and consolidate data, distill feedback into digestible chunks, and improve my ability to execute with polish, precision, and efficiency. My ideas, my problem-solving abilities, and my creative vision are still entirely mine — AI is, quite simply, a tool I can use to help me get to a shippable UX solution faster and with fewer roadblocks.

Where AI Fits into My Process

AI plays a role in several key areas of my professional workflow:

• Writing & Communication – I tend toward over-explaining and using hyper-precise language, which can lead to information overload and bury the key points. AI helps me distill my ideas, keeping structure tight and focus clear without sacrificing nuance or specificity.
• Design Quality – Checking accessibility compliance, catching visual inconsistencies, and verifying adherence to style guidelines — often spotting issues that are tedious for a human eyes to discern on-screen. For example, during AWS branding updates, AI helped identify subtle color, font, and component mismatches that would have been time-consuming and error-prone to track down manually.
• Organizational Support – Turning my existing backlog and time estimates into sprint plans, prioritizing based on constraints, and keeping workstreams and reporting on track. I keep inputs targeted and consistent so the output stays relevant and actionable.

Principles for Responsible Use

I follow a “trust but verify” approach. AI can construct a persuasive argument, but it’s not always correct, so I treat every output as a suggestion, never a directive, and I always fact check anything I have an inkling of doubt over.

• Critical Evaluation – If AI produces something that seems questionable, I press for more detail until I’m confident it’s accurate. If I can’t find a clear answer I’ll set it aside for further investigation or just outright discard it.
• Maintain Ownership – I only present work I can fully stand behind and defend. Anything that doesn’t align with my professional judgment gets reworked or removed.
• Voice and Authenticity – Especially in writing, I ensure the final product still reflects my perspective and voice. Anything less risks feeling inauthentic or leaving me unable to speak to it confidently.
• Clear Boundaries – When AI is used for expressly creative outputs (like placeholder imagery or conceptual mockups), I make that clear — especially when they’re intended only to communicate a concept rather than serve as final deliverables.

Why I Don’t Always Disclose AI Involvement

I’m transparent about using AI in general, but I don’t label every deliverable that AI touched. That’s because I use it to enhance my own ideas, not substitute for them.

Attribution makes sense when AI generates a creative element wholesale. But when I’m directing the work, shaping the content, and making final decisions, the authorship — and responsibility — is mine. Over-attributing AI in these cases risks undermining the very expertise and rigor that make my work effective.

The Future of AI

In the next 3–5 years, I expect AI will be able to produce passable UX from a simple prompt, especially when trained with a robust design system. For some cases, that will be “good enough.”

And that is fine.

But great UX isn’t just about following the rules or assembling components — it’s about uncovering the root cause behind a problem, not just addressing the symptoms. That requires perspective, empathy, and nuanced reasoning that AI cannot yet replicate.

People can process innumerable intricacies of human nature in an instant and without conscious effort: subtle differences in inflection and language, small gestures or facial expressions, and contextual behaviors and mannerisms. These cues all enhance how we empathize with a customer. AI, while powerful, cannot yet do this — and because its outputs are only as good as its inputs, it still has a long way to go in capturing the full context required to truly empathize with another human.

I’ll continue to embrace AI as a tool to enhance speed, accuracy, and quality — but never at the cost of the human insight and care that turn functional experiences into truly exceptional ones.

What do you think?

This is my take on where AI shines, where it still falls short, and how I try to use it responsibly without losing the human touch. Maybe I’m right? Maybe I’m wrong? Maybe I’m just a designer who’s had too much coffee (likely) and too many conversations with chatbots (no comment).

How do you approach AI in your work? Do you avoid it or let it run wild? I’d love to hear how others are navigating this in their daily lives.