Safer choices start with better judgment.
Young people spend hours a day reading health and lifestyle content online — but only 26% feel confident judging what's actually true. VITAL helps them tell the difference, right where they're already reading.
The brief came from MediaFutures, a Bergen research center working on responsible AI-driven media technology. From there, our team of five ran the full process — research, design, and a working prototype.
10 interviews (8 target users 17–25, 2 experts), 3 focus groups by age, and document analysis — followed by 5 rounds of prototype user testing.
Trust in AI splits by age — younger users distrust it outright; older users are open to it, but only if it's transparent about its sources and confidence. Everyone agreed on one thing: health is higher-risk, so AI needs to show its work.
From there: a co-design workshop with upper-secondary students, a design sprint (Crazy 8s, storyboards, prioritization), then low-fidelity sketches built up to the working prototype.
Testing made one thing clear: a confident AI verdict on health feels risky. So VITAL avoids the usual red/green certainty — claims are shown as light blue (not fully true) or orange (not fully wrong), with sources always visible, keeping the reader in the loop rather than handing them a verdict.
A working Chrome extension prototype, tested across 5 rounds with real users. Color-coding and in-text highlighting gave people fast, clear understanding — confirmed in every round.