lesson
20 minAI in the Design Process
Where AI tools help design work, where they do not, and the risks to watch for.
What you will learn in this lesson
You will decide when an AI tool helps your design work and when it puts people at risk.
What AI can do in design
AI tools can now draft screen text, suggest layouts, make placeholder images, and summarize notes. Some design tools have AI features built in. Used well, they save time on first drafts.
Where AI helps, and where it doesn't
| Use AI for | Don't use AI for |
|---|---|
| First drafts of button labels and help text | Deciding what users need |
| A list of edge cases to consider | Replacing interviews or usability tests |
| Rewriting text at a lower reading level | Making fake quotes or fake research |
| Translating, as a starting draft | Final translations without a fluent human check |
| Brainstorming layout options | Images of "real" customers or community members |
The risks
- Made-up research. AI can invent quotes and data that sound real. Using them in a case study is dishonest. It also leads to bad designs.
- Bias. AI learned from the internet. It can repeat stereotypes in personas, images, and text.
- Privacy. What you paste into an AI tool may be stored. Never paste names, contact details, or anything about minors.
- Sameness. AI layouts look like everything else. Your research is what makes your design fit your users.
Practice: make your screen text clearer
Use AI to rewrite help text at a lower reading level. Then read it out loud. Check that every fact is still true.
Prompt Lab
Rewrite one screen's text in plain language. Then check every fact yourself.
Your prompt
Act as a plain-language editor who writes at a 6th-grade reading level. Format: The rewritten text, then a list of anything you changed that might change the meaning