AI Prompts for UX Designers gives US UX designers, product designers, and UX researchers copy-paste prompts for the writing that surrounds the design work — research plans, usability test scripts, persona synthesis, accessibility audit summaries, and the case study copy you need for your portfolio. These are for the parts of design where a well-written doc unlocks weeks of downstream work.
Every prompt is written for how design actually happens in US product teams: research plans that survive stakeholder pushback, usability scripts that produce real signal, personas built from real interview data (not marketing archetypes), and handoff docs engineers can actually build from. Fill in the brackets with your real project context, then edit the output against what you know about your users and your team.
Do not paste raw research recordings, participant names, unblurred screens from a production app under NDA, or any PII from user studies into a public AI tool. For research synthesis with real participant data, use an enterprise AI, your research repository's built-in AI features, or scrub the transcript before pasting.
AI Prompts for UX Designers gives US UX designers, product designers, and UX researchers copy-paste prompts for the writing that surrounds the design work — research plans, usability test scripts, persona synthesis, accessibility audit summaries, and the case study copy you need for your portfolio. These are for the parts of design where a well-written doc unlocks weeks of downstream work.
Every prompt is written for how design actually happens in US product teams: research plans that survive stakeholder pushback, usability scripts that produce real signal, personas built from real interview data (not marketing archetypes), and handoff docs engineers can actually build from. Fill in the brackets with your real project context, then edit the output against what you know about your users and your team.
Do not paste raw research recordings, participant names, unblurred screens from a production app under NDA, or any PII from user studies into a public AI tool. For research synthesis with real participant data, use an enterprise AI, your research repository's built-in AI features, or scrub the transcript before pasting.
Guides, tips, and deep dives for this prompt category
The best free AI prompt library in 2026. Browse 4,000+ tested prompts for Midjourney, DALL-E 3, Stable Diffusion, FLUX, and ChatGPT. No signup, no paywall — copy and create instantly.
Read moreAI GuidesMaster FLUX AI with this complete guide. Learn FLUX.1 Pro vs Dev vs Schnell, prompt techniques, and get 50 ready-to-use FLUX prompts.
Read moreCopy any prompt below, paste into ChatGPT, Claude, Gemini, or Copilot, and fill in the placeholders in [brackets].
Act as a US senior UX researcher. Write a research plan for [project name]. Sections: Background (why now, 2 sentences), Research question(s) (max 3, phrased as questions not topics), Method and rationale (why this method for this question), Participants (segment, sample size, recruit criteria), Timeline, Deliverables, Out of scope, Stakeholder review points. Assume a skeptical PM will read this and ask "why not just look at the analytics?" — anticipate that.
Act as a US UX researcher. Write a moderated usability test script for [product/feature]. Include: warm-up questions (2), 4 to 6 tasks phrased as user goals (not clicks — "find out how much you'd pay this month" not "click the pricing page"), think-aloud prompts, post-task questions, and wrap-up. Total session: 45 minutes. Add a note on what NOT to say (leading questions, feature justifications). Neutral, curious moderator tone.
Act as a US UX researcher. Build a persona from this interview data: [paste anonymized notes and quotes from 5 to 8 interviews]. Output: role and context (2 sentences), primary goals (3, in the user's language), key frustrations with the current process (with a supporting quote for each), tools they use today, one behavior that surprised us, and the one thing that would make their job easier this week. Only use what's in the pasted data — do not invent details.
Act as a US accessibility specialist. Write an accessibility audit summary for [product/section]. Structure: Scope (what was audited, what was not), Standard (WCAG 2.1 AA), Findings by severity (critical / serious / moderate / minor) with WCAG success criterion referenced for each, screenshot placeholders, recommended remediation, and estimated effort. Written for a product manager to prioritize — not just a list of violations.
Act as a US senior product designer running a design critique. Write a critique feedback structure for reviewing [design work — e.g., a new onboarding flow]. Include: framing the critique (what feedback is useful, what isn't, 2 sentences), lens 1 (does the design solve the user problem — 3 questions), lens 2 (does the design fit the system — 3 questions), lens 3 (what's the strongest and weakest part), and a note on how to give feedback that's specific and actionable rather than preference-based.
Act as a US product designer. Write a portfolio case study outline for [project]. Sections: One-line summary (role, company, outcome), Context and constraints, The problem I was solving, Research and what I learned, Design decisions with the tradeoff for each, What shipped, Measured impact (be honest — if not measured, say so), and What I'd do differently. Target audience: hiring managers at [type of company]. Include a "what I got wrong" paragraph — that's the section that differentiates.
Act as a US UX researcher facilitating a workshop. Write a facilitation guide for a [type of workshop — e.g., stakeholder alignment, journey mapping, design sprint] session with [number] participants over [duration]. Include: pre-work for attendees, opening framing (5 min), activities with timeboxes and instructions, breaks, decision points, and how to end with clear next steps. Add a section on what to do if the room goes off track.
Act as a US UX researcher. Write stakeholder interview questions for [project — e.g., discovery for a new product area]. 8 to 10 questions total, mix of context questions (their team, their goals), problem questions (what they see users struggling with), and constraint questions (what's already been tried, what's off the table). Open-ended, no yes/no questions. Include a closing question about who else I should talk to.
Act as a US UX researcher. Write a journey map data collection plan for [user journey — e.g., first-time onboarding, checkout, renewal]. Cover: data sources (interviews, analytics, support tickets, session recordings), what to capture at each stage (actions, thoughts, emotions, pain points, opportunities), how many participants or data points per stage, and how to reconcile conflicting signals across sources.
Act as a US product designer. Write wireframe annotation notes for [feature] handoff. For each key screen, include: purpose in one sentence, user state (first-time / returning / error state), interactions (what happens on click, hover, keyboard focus), copy notes with source (final vs. placeholder), accessibility notes (focus order, ARIA labels needed, contrast ratios), and open questions with an owner. Written so an engineer can build without a meeting.
Act as a US design systems designer. Write a component documentation page for [component name — e.g., Button, Modal, Toast]. Sections: What it is (1 sentence), When to use (3 bullets), When not to use (2 bullets — this is the hard one), Anatomy with labeled parts, Props/variants with visual examples, Accessibility requirements, Do and Don't examples, Related components. Written for both designers and engineers using the same doc.
Act as a US product designer. Write a design handoff to engineering checklist for [feature]. Cover: linked Figma file with named frames, design tokens used, states covered (default / hover / focus / active / disabled / loading / empty / error), responsive breakpoints, accessibility notes (keyboard, screen reader, contrast), copy source of truth, motion specs if any, edge cases documented, open questions with owners, and a review meeting recap link placeholder.
Act as a US product designer. Write an A/B test hypothesis for [design change] on [surface]. Format: We believe that [specific change] for [specific user segment] will result in [specific measurable outcome] because [insight or evidence]. We'll know we're right when [primary metric] moves by [threshold] within [time window], and we'll watch [guardrail metric] to make sure we don't regress [risk]. Include what you'd learn if the hypothesis is wrong.
Act as a US UX research operations specialist. Write a user interview recruitment screener for [target participant profile]. Include: 6 to 10 screening questions with the correct answers marked, disqualifying criteria (competitors, prior participants in last 90 days), demographic questions relevant to the study (not gratuitous ones), incentive amount placeholder, time commitment, and consent language. Screener should take under 3 minutes to complete.
Understanding the building blocks lets you adapt any prompt to your own creative direction.
Tell the AI who the output is for and what real workplace situation it should support.
Act as a federal program analyst preparing a plain-language memo for agency leadership.Name the exact deliverable: email, memo, checklist, SOP, meeting recap, training note, or status update.
Format the answer as a one-page briefing with bullets, risks, and next actions.Specify whether the output should sound official, executive-ready, plain-language, or employee-friendly.
Use a professional, neutral, public-sector tone suitable for a US agency audience.For government, HR, finance, healthcare, legal, and compliance workflows, accuracy guardrails matter more than clever wording.
Use only the facts below, flag assumptions, and include a section for items that need verification.Ask the model to surface uncertainty so the user can verify sensitive or official information before using it.
Before finalizing, list compliance risks, missing details, and any claims that need human review.Tested on this prompt category as of mid-2026. Ratings reflect quality for AI Prompts for UX Designers specifically.
| Model | Best for | Rating |
|---|---|---|
| ChatGPT (GPT-4o / GPT-5) | Everyday drafting and summaries | |
| Claude Sonnet 4.5 | Long documents and policy | |
| Gemini 2.5 Pro | Grounded in Google workspace | |
| Copilot (M365) | Office 365 integration | |
| Perplexity | Answers with citations |
Ratings reflect suitability for this category. Free tiers available on all listed models. Last tested May 2026 by PromptSpace editors.
Yes for structure, no for defensibility. AI produces a solid plan template — background, question, method, sample, timeline. What it can't do is anticipate your specific stakeholders' pushback ("why not just look at the analytics?"). Add one sentence to the prompt naming the objection you expect, and the AI folds a response into the plan.
Strip identifying details before pasting — participant codes instead of names, role descriptors instead of employer names, anonymized quotes. For any research repository or enterprise research tool, use its built-in AI (which stays within your data boundary) rather than pasting into a public chat. Your research ops or legal team can tell you what's cleared.
Only when built from real data you paste in. AI generating a persona from scratch produces a marketing archetype that any experienced PM will recognize as fiction. AI synthesizing a persona from 8 real interview quotes you paste in produces a grounded artifact you can defend. The difference is entirely in what you feed it.
It helps with the structural scaffolding and the honesty section that most designers avoid writing. Ask specifically for a "what I got wrong" or "constraint I couldn't solve" paragraph — that's the section hiring managers remember. AI will draft it if you feed it a real constraint from the project; don't let it invent one.
AI can format findings against WCAG success criteria and write remediation notes if you paste in the specific issues (screenshots, DOM snippets, screen reader output). It cannot detect issues from a screenshot alone reliably — pair it with axe DevTools, WAVE, or a manual keyboard/screen-reader pass, then have AI structure the report.
Learn the basics of creating stunning AI-generated images using prompts from our library.
GuideDiscover the secrets to crafting prompts that produce consistent, high-quality results.
CollectionCopy-paste 100 tested Midjourney v6 prompts: portraits, cinematic, fantasy, product shots & more. Free, updated for 2026 - instant results.
Social MediaCreate scroll-stopping Instagram content with these AI image prompts designed for Reels, Stories, and posts.
Browse our full library of ai prompts for ux designers — all free, copy-paste ready, no signup.
Or use our AI Prompt Generator to create custom prompts for your exact style in seconds.
Start with the artifact that will unblock the most people. If eng is waiting on handoff, run the handoff checklist prompt. If a stakeholder is pushing back on your research approach, run the research plan prompt to produce a doc that answers the objection in writing. Fill in the placeholders — project name, target users, research question, ship date — with real specifics.
For any research or synthesis prompt, feed the AI real inputs. "Synthesize a persona from user data" without the data produces marketing fiction. Paste anonymized interview quotes, behavioral data, or observation notes and the output becomes a persona grounded in actual research rather than an archetype you'll have to defend later.
Never paste participant names, employer names, or any detail from a research session that could identify someone. This applies even inside enterprise AI unless your legal team has explicitly cleared it. Use participant codes (P01, P02) or role descriptors ("mid-market ops lead, ~2 years tenure") and strip identifying phrases from quotes before synthesis.
Do not use AI to generate research findings from data you haven't collected. AI will happily invent plausible insights, personas, or usability issues if you ask it to. If the prompt requires user data, paste real data. If you're at the plan stage before research, use AI for the plan structure, not for pretending you already ran the study.
For startup design teams (design team of 1 to 3), skip the formal deliverables. Add "we are a scrappy design team — no formal design review process, output is Figma + a Loom" to any prompt and the AI cuts the enterprise scaffolding. For portfolio case studies as a working designer, specify the audience (hiring managers at consumer apps, at B2B SaaS, or at design agencies) — the framing shifts significantly.
For enterprise UX (banks, healthcare, government), specify the constraints upfront: WCAG 2.1 AA compliance target, regulated industry, multi-stakeholder review process, existing design system constraints. The AI produces docs that anticipate the review questions instead of docs that will get flagged on the first pass.
The fastest tell of AI-drafted UX writing: generic personas ("Sarah, 34, marketing manager, values efficiency"). Real personas from real research have specific verbatims, actual workflow details, and a named tension the person navigates every day. Replace every abstract trait with a behavior you observed in a session and the persona reads as research output, not marketing filler.
For portfolio case studies, AI drafts default to a linear problem-process-solution structure that every designer has already seen. The differentiator is the honesty section — what you tried that didn't work, what you'd change, what constraint shaped the final call. Ask the AI for a case study with a "what I got wrong" paragraph and it produces a more interesting artifact than a polished linear narrative.
Yes for structure, no for defensibility. AI produces a solid plan template — background, question, method, sample, timeline. What it can't do is anticipate your specific stakeholders' pushback ("why not just look at the analytics?"). Add one sentence to the prompt naming the objection you expect, and the AI folds a response into the plan.
Strip identifying details before pasting — participant codes instead of names, role descriptors instead of employer names, anonymized quotes. For any research repository or enterprise research tool, use its built-in AI (which stays within your data boundary) rather than pasting into a public chat. Your research ops or legal team can tell you what's cleared.
Only when built from real data you paste in. AI generating a persona from scratch produces a marketing archetype that any experienced PM will recognize as fiction. AI synthesizing a persona from 8 real interview quotes you paste in produces a grounded artifact you can defend. The difference is entirely in what you feed it.
It helps with the structural scaffolding and the honesty section that most designers avoid writing. Ask specifically for a "what I got wrong" or "constraint I couldn't solve" paragraph — that's the section hiring managers remember. AI will draft it if you feed it a real constraint from the project; don't let it invent one.
AI can format findings against WCAG success criteria and write remediation notes if you paste in the specific issues (screenshots, DOM snippets, screen reader output). It cannot detect issues from a screenshot alone reliably — pair it with axe DevTools, WAVE, or a manual keyboard/screen-reader pass, then have AI structure the report.