How to Champion Human-Centered Design in the Age of AI

In my last AI-focused post, I made the case that AI isn’t something designers need to fear, it’s something we’re built to lead. I laid out three principles to reframe the conversation and keep our work deeply human, even as the tools around us get faster, smarter, and more automated.

Once we’ve stopped defending, we’re ready to start building. Once the fear settles, the real opportunity comes into view.

So let’s talk about what it looks like to practice those principles. Let’s talk about how to champion human-centered design with AI... not in spite of it.

Let AI Surface Patterns. Let People Make Them Matter.

One of AI’s greatest strengths is speed. It can churn through data, cluster feedback, and identify behavioral trends at a pace no human can match.

But fast ≠ insightful.

Human-centered design isn’t just about finding the signal. It’s about asking:

• What does this mean for the people we serve?

• What’s the emotional undercurrent?

• What’s driving the behavior?

This 👆 is where designers shine.

AI can show you "what" is happening, but it takes human judgment to ask "why" and whether or not that “why” is something we should lean into or solve for.

Try this:

• Use AI to analyze support logs, tag themes in research, or highlight where users drop off.

• Then slow down. Talk with your team. What’s obvious? What’s nuanced? What’s being said between the lines?

  
  

Real-world example:

UXtweak used AI to generate follow-up questions during usability testing, helping researchers move faster. But it was the humans who decided which questions led to meaningful insight. AI suggested. Humans interpreted.

Let AI Draft. You Direct.

As I wrote previously “Human-centered design will always require humans.” An important way that shows up is in how we harness and aim AI's speed and balance it with smart, human-centered decision-making.

AI can produce first drafts in seconds—flows, UI concepts, microcopy, even research summaries. That’s not something to fear. It’s a head start. Great design isn’t defined by how fast you move, it’s defined by what you choose to keep, what you refine, and what you don’t ship.

Try this:

• Ask AI to generate a first pass. Then step in as the editor, not the executor.

• Shape the tone. Clarify the message. Anchor it to strategy, not novelty.

Real-world example:

In a recent study, researchers tested how well GPT-4 could perform UX evaluations on mobile app screenshots. The AI was able to identify common design issues like low contrast, cluttered layouts, and inconsistent touch targets faster than any human could.

But that’s where the automation stopped.

The AI critiques were often vague, lacked prioritization, and missed contextual cues. Designers were still essential to interpret which issues truly impacted the user experience and how to fix them in a way that aligned with goals, brand, and real-world use cases.

Start with the Problem, Not the Prompt

One of the most common mistakes I see is when design teams start with what the AI can do, instead of what the user actually needs. Just because you can generate something doesn’t mean it’s worth generating. Good design still starts with context, constraints, and clarity.

Try this:

• Before writing a prompt, write a problem statement.

• Ask: What’s the core need here? Who’s affected? What’s the outcome we’re aiming for?

• Then bring AI in—as a partner, not a compass.

Real-world example:

In a study of adolescents using AI-powered mental health tools, researchers found that teens didn’t want fully automated advice. Instead, they wanted a sense of partnership and control. The takeaway? If you frame the problem as “How can we automate support?” you get one kind of solution. But if you ask, “How do we help teens feel safe, seen, and supported?” you build something that works.

Make AI a Team Tool, Not a Designer Shortcut

This is something I didn’t explore much in the last AI focused post, but it’s crucial. Designers aren’t the only ones who benefit from AI. When you bring other disciplines into the loop, things get better for everyone. From research to ops to engineering, cross-functional teams can use AI to align faster, generate alternatives, and synthesize more clearly.

Try this:

• Use AI to prep for working sessions, map journey touchpoints, or generate what-if scenarios.

• Review the outputs together. Let different perspectives shape the next steps.

  
  

Real-world example:

At Lightful, a cross-functional "AI Squad" composed of designers, product managers, and engineers used AI tools in tandem to prototype faster, test more ideas, and align on decisions together each day. AI generated multiple interface options and microcopy variations. The team then reviewed them as a group (deprioritizing some, refining others) before deciding which to move forward with. This approach accelerated alignment and created shared ownership over AI-assisted outputs.

Let AI Handle the Volume. You Stay Close to the Meaning.

AI can scale your reach. But only you can stay close to what matters. It’s easy to drown in data or offload every insight-sorting task to a model. But when we lose the human thread (e.g.; curiosity, nuance, care) we lose the very thing that makes design worth doing.

Try this:

• Let AI process the bulk. Let humans connect the dots.

• Make time to sit with a transcript, listen to a frustrated user, or read a customer’s words.

• Use AI to accelerate. Use your humanity to prioritize.

Real-world example:

According to a Wall Street Journal report, companies are increasingly using AI to manage customer interactions... but users aren’t sold. In one survey, 77% said they preferred talking to a human, and over half reported frustration after engaging with a chatbot (even when their issue got resolved).The lesson? AI can handle the workload. But when things get personal, nuanced, or emotionally charged, meaning still matters. People still prefer people.

AI can manage the volume, but you need to stay close to the moments that actually move users.

The Work Ahead

This isn’t about resisting AI. It’s about guiding it. If you’ve already shifted your mindset (via my Anxiety to Advantage post), this post is your next step. Not a list of tools. A blueprint for action.

The teams that win with AI won’t be the ones who move the fastest. They’ll be the ones who stay focused. Grounded. Human.

Let’s build with both hands. Let’s stay curious. Let’s stay human.