Hey, it’s Fiona
When it comes to AI, trust isn’t won by features. It’s won by transparency.
I was reminded of this the other day when I asked an AI shopping assistant to “find me organic cotton T-shirts under £40.” It gave me five results, but no clue why those five. Were they the cheapest? The highest quality? The most popular? With no context, I didn’t trust the list and went straight to another site to check instead.
That’s the danger. When AI outputs feel like a black box, people assume the worst. It doesn’t matter how advanced the model is… if the experience feels unfair or unpredictable, trust evaporates.
The three lenses of AI trust
Designing for transparency really comes down to three things.
The first is explainability. People need to know why a decision was made. Imagine a credit scoring tool. If it says “declined because income falls below the threshold,” that feels fairer than a flat “rejected.” One gives you context, the other leaves you guessing.
The second is control. AI should act with the user, not at them. A shopping assistant that recommends a “best pick” but still lets you adjust filters creates agency. A news app that hides its sources does the opposite, leaving you powerless.
And the third is reversibility. Machines make mistakes, just like humans. What matters is whether you can recover. A scheduling app that suggests a time but lets you instantly reschedule feels safe. A healthcare portal that updates records automatically without confirmation feels anything but.
Miss any one of these three, and transparency starts to crack.
What transparency looks like in practice
The shape of transparency changes across industries, but the principle is always the same: explain the why. For example:
A SaaS dashboard might highlight the three metrics behind a risk score.
A marketplace could disclose why a provider is “recommended,” whether that’s response time or ratings.
An online store might reveal that its “best match” is ranked by price, materials, and delivery time.
Even a content site earns trust simply by attaching a byline and a date.
These small touches are not decorative extras. They’re the difference between “I trust this result” and “I don’t believe it.”
Opaque design carries a cost. People second-guess results, adoption slows, and in some sectors, regulators step in. Finance, healthcare, and hiring are already in the spotlight.
Think of transparency as insurance. It may feel like extra effort now, but it prevents the kind of blow-ups that sink products later.
Why this isn’t just about humans
Interestingly, transparency isn’t only for people. Other AI agents are reading your product too. If they can’t parse your reasoning, they may not trust or amplify it.
For humans, reasoning must be readable.
For agents, it must be machine-parsable — structured data, schema, and factual labels.
Dual-layer clarity is what keeps you both visible and credible.
Try this this week
Pick one AI-powered feature in your product and stress-test it. If the result surprised a user, could they see why? Could they shape or override it? And if it was wrong, could they undo it?
Any “no” is a transparency gap waiting to be fixed.
Why I’m excited about this
It’s tempting to see transparency as a compliance box to tick. But I think it’s one of the most exciting design challenges we face right now. Just as mobile design forced us to rethink layout, transparency is pushing us to rethink communication. Done well, it can set you apart. Competitors may look shinier, but if your product feels more trustworthy, you’ll win the long game.
Until next week, I’d love to hear your thoughts:
Where have you seen AI transparency done well — or badly?
Hit reply and let me know.
Talk soon,
Fiona
Fiona Burns
Work with me
Alongside writing Beyond the Screen, I help founders and product teams design digital products their users (and AI agents) can’t ignore.
That might mean validating an early idea, shaping the first version of a marketplace, or redesigning a website so it’s easier for both people and machines to understand.
If you’re building something new and need UX/UI support, head over to my website to see how we could work together.