Fifteen years ago, I studied AI. I built neural networks, worked with robots, and was excited by the idea of making systems smarter. But I eventually found myself asking a different question: why make things smarter if we are not making them better? That question led me into UX research.

Today, AI is at the centre of product development. It helps us write, analyse, ideate, summarise, and move faster. But it is also starting to play a more questionable role: teams are asking AI to validate product ideas, simulate user feedback, create personas, and tell them what users need.

This talk is about how to use AI well in UX research and where to draw the line.

AI can make UX research faster, clearer, and easier to share. Used well, it can support the work around research and help teams think more sharply. But it cannot replace what matters most: contact with real users. It cannot observe real behaviour, understand context, or reveal the messy reality of how people actually use products.

AI can make your ideas sound convincing. UX research tells you whether they are true.