AI and the New Era of UX Research: Why Human Judgement Matters More Than Ever

AI can generate answers quickly. Our enduring skill is knowing which questions are worth asking, which answers can be trusted, and when the answer must come from a real human being.

Artificial intelligence is changing UX research faster than expected. After more than a decade working as a researcher across academia and consulting, I have watched research evolve into a strategic part of product development. Industry data shows that 80% of researchers now use AI in their research workflows (User Interviews, 2025). AI tools now transcribe interviews, cluster themes, and draft first-pass summaries in minutes. Work that used to take days of manual coding is compressing into hours, freeing researchers for the part of the job that actually needs judgment: deciding what a pattern means for a decision a stakeholder has to make tomorrow. Yet this shift is not simply operational; it is fundamentally changing what we value as researchers.

One of the most debated developments is the rise of “synthetic users”: AI-generated representations of users that can simulate interviews, explore hypotheses and test assumptions (Sullivan, 2026). Yet practitioners remain cautious. In User Interviews’ 2026 State of Synthetic Users report, 47% of researchers described themselves as sceptical and wanting more evidence before going further. I consider myself one of them. As UXer Prajna Hegde notes, synthetic users are not a magic wand; researchers first need to clarify what problem they are actually trying to solve (Hegde, 2026). Designer Alina Bezchotnikova reinforces this distinction, arguing that synthetic users should not be presented as “digital twins” (borrows unearned engineering credibility): models can simulate responses, but they do not reproduce the complexity of lived human experience (Bezchotnikova, 2026).

This distinction is critical. Speed is not understanding. AI can process vast amounts of information and surface patterns quickly, but it can also reproduce hidden biases and flatten the contextual nuances that an experienced researcher would notice when investigating a problem, service or product experience. Maria Rosala emphasises that even when AI can produce research-like outputs, outsourcing the learning process risks losing the empathy, shared understanding and conviction that come from directly observing and engaging with users (Rosala, 2026).

As Emmanuelle Savarit explores in her podcast What Happens to UX Research Craft in the AI Era?, AI is not necessarily eliminating research craft; it is forcing us to reconsider what that craft actually consists of. If we define research primarily through the production of artefacts (transcripts, reports, user needs and presentations) then automation can make our work appear increasingly commoditised. If we define it through critical thinking, methodological rigour, interpretation and responsible decision-making, the role becomes much harder to automate (Savarit, 2026).

The market is already moving towards researchers who can work alongside AI while maintaining methodological integrity and strategic judgement (Venkat, 2026).

The future of UXR is therefore not simply about competing with AI on speed. It is about knowing where speed creates value and where it creates risk. Our role is increasingly to apply critical thinking and ethical oversight to design trustworthy systems of inquiry; to decide what can be automated, what needs human judgement, and whose experiences still need to be heard.

AI can generate answers quickly. Our enduring skill is knowing which questions are worth asking, which answers can be trusted, and when the answer must come from a real human being.

References

• Bezchotnikova, A. (2026). Synthetic users are not digital twins. Medium Bootcamp. synthetic-users-are-not-digital-twins
• Hegde, P. (2026). Synthetic users, Chapter 1. Medium. synthetic-users-chapter-1
• Rosala, M. (2026). Don’t outsource the learning: Why human-led research still matters in the age of AI. Nielsen Norman Group. human-led-research-still-matters
• Savarit, E. (2026). What Happens to UX Research Craft in the AI Era? [Podcast Episode]. Spotify.
• Sullivan, C. (2026). Your synthetic users primer. AI Customer Research. your-synthetic-users-primer
• User Interviews. (2025). The 2025 State of User Research Report. data-reports
• User Interviews. (2026). The State of Synthetic Users Report. state-of-synthetic-users-report
• Venkat, J. (2026). The 2026 UX research job description: What AI frontier companies want now. Substack. the-2026-ux-research-job-description