When AI Enters the Conversation: Rethinking the Designer’s Ethical Role

What are the ethical responsibilities of interaction designers to their end users? Especially related to AI and its integration with our work.

When we talk about design, we don’t usually talk about moral responsibility. We tend to focus on what a design can accomplish or how it can improve someone’s life. After all, design exists to serve people. The intention behind our work is rooted in human benefit.

But once AI becomes part of the picture, our first reaction often shifts toward ethics. When AI is used as a design tool, we can no longer fully control what the final product will be or how it might influence users. AI isn’t completely predictable. Take language models as an example: they inherit the biases of the internet, and they don’t understand right or wrong. They generate answers by following patterns. Sometimes, even I catch myself believing something incorrect simply because the AI presents it in such a convincing way.

That’s why, as designers—especially interaction designers, who often design systems involving human–AI interaction—we carry specific ethical responsibilities.

First, we need to make sure users understand what the AI is doing.
What parts are AI-generated? What parts are not? How does the system work? This is basic but essential. Before I understood how AI worked, I trusted its responses too easily.

Second, we must protect privacy and personal data.
This one is self-explanatory but non-negotiable.

Third, we have to minimize bias and avoid manipulation.
Whether we are using AI or designing new AI-driven experiences, we should resist relying on biased or non-objective information. These distortions can mislead users and create harm on a social level. Even if bias can’t be completely eliminated, we need to stay aware of its presence and actively counteract it.

As AI becomes more deeply woven into interaction design, our responsibility isn’t just to build useful tools—it’s to build responsible ones. Our choices influence how people understand the world, and that influence deserves care.

(This blog post was translated and revised with the assistance of ChatGPT.)