UX survey respondents flag AI quality and privacy barriers

What the survey establishes

Published October 23, the survey ran August 24–September 11, 2026. Recruitment used NN/G’s panel, newsletter, Slack communities and LinkedIn. From 2,814 initial responses, the researchers excluded bots, incomplete responses and practitioners not currently working, retaining 521.

This convenience sample is not representative of all UX workers. The percentages describe reported barriers; they do not measure productivity or employment effects.

Sources: Nielsen Norman Group

Our interpretation: try a small review exercise

Imagine a researcher preparing a presentation about an invented shopping app. Give an AI tool five fictional interview notes and ask for a one-page summary. Before running the task, write down what an acceptable result must contain: each finding must link to a note, disagreements must remain visible, and unsupported claims must be marked as questions.

Have a colleague review the result against those criteria. Record which sentences they accepted, changed or removed, and why. This creates something concrete to discuss when deciding whether the tool belongs in that particular workflow. It also makes a polished but unsupported conclusion easier to challenge.

Keep the exercise separate from a decision to upload real interviews. For an actual project, identify who can approve the proposed use of participant material and what information they need to decide. An exercise with invented notes cannot settle that question.

For a portfolio, the useful artifact could be a short review memo showing one rejected claim and the evidence behind that decision. We suggest this as a way to explain professional judgment, not as proof of a hiring advantage or a conclusion drawn by the survey’s authors.