Working on this project and in the AI in Design course made me much more intentional about when AI actually belongs in my process. I treat AI like any other tool, it needs to either improve how I work or improve the quality of the final output. If it does not do either, then using it adds ethical and environmental cost without a good reason, and that alone is enough for me to pause.

I use AI most often as a research partner, a synthesis helper, and a way to check for risks and blind spots. Coming from a business and marketing background, I know I am prone to confirmation bias, especially when I am working alone. Inviting AI into the process gives me alternative interpretations of interviews, surfaces patterns I may have missed, and forces me to look at risks that do not fit my original narrative. I still start by building my own structure and point of view, then ask AI to expand, question, or pressure test it rather than generate ideas from scratch.

Over time, I have developed a set of criteria that I apply whenever I reach for these tools. I ask whether my own thinking is still central, whether AI is reducing bias instead of amplifying it, and whether it is introducing any new ethical concerns such as privacy issues or misrepresentation. I also pay attention to places where AI makes things look more polished than they really are, especially in prototypes, since that can distract users from the core idea we are trying to test. Going forward, I want to be transparent about where AI shows up in my work and hold it to the same standard as any other design material. It should make the work more thoughtful, more rigorous, and more grounded in reality, not just faster.