Kavi, the KAVI owl, waving on a book
HomeAcademyAIWho it's forPricingFAQLabsContactGet Started
AI Literacy · Stage 6

Bias, Fairness, Provenance

A smooth AI answer can still be one-sided, and it can come from nowhere you can check. Learn to evaluate output for bias and fairness — and to trace where it really came from.

✦ AI Literacy📘 Stage 6 · S6⏱ ~9 min🧒 Ages 10–12★ Free to start
I can evaluate output for bias and check provenance.
Start the free lesson →⬇ Printable PDF
Kavi the owl for AI Literacy
📖

Kavi Traces a Too-Perfect Photo

A helper shows Kavi a photo of "a scientist" — and every single time, it's the same kind of man. "Where is everyone else?" Kavi wonders. "Scientists are women and men, from all over the world." Then Kavi spots a six-fingered hand in the picture and a 'source' link that leads nowhere. So Kavi does two things: checks if the output is fair (it isn't — it's one-sided and biased), and traces where it came from (no author, a dead link — weak provenance). A smooth answer can still be biased and come from nowhere you can check.

🎓

What your child will learn

1One-sided vs balancedAI learns from the examples it is shown, so its output can pick up one-sided patterns. Tap each card!
2Biased, or where's it from?Look at each output. Decide, then tap to check. Tap Next test for more.
3Good provenance check — or not?Which moves really help you trace where something came from? Tap a bin for each one!
🔍

A peek inside

An AI image tool shows 'a doctor' as a man in every single picture. How do I evaluate this for bias and fairness?
I ask the fairness question: does this show many kinds of people, or the same one over and over? Every picture is one kind of person, so whole groups — women, and people of many ages and backgrounds — are left out. That is a <b>one-sided, biased</b> pattern. A fair tool would show doctors who are women and men, of many skin tones and ages.

Try this at home: What does it mean to say an AI output is <b>biased</b>?

💬

Questions parents ask

+What does AI bias mean?

AI bias is when an output keeps showing only one kind of person or one side — like always drawing 'a scientist' as the same man — so whole groups get left out. It learns these one-sided patterns from the examples it was trained on.

+How do you check if an AI output is fair?

Ask whether it shows many kinds of people and more than one point of view, instead of the same single kind over and over. A balanced output is fair; a one-sided one is biased.

+What is provenance, and how do you check it?

Provenance is where a piece of content came from and who made it. You check it by tracing the author, finding a real source you can open, looking at the date and place, and seeing if a trusted source agrees — not by counting likes or trusting a confident tone.

+Why can't lots of likes or shares prove something is true?

Popularity is not provenance. A huge number of shares still can't tell you who first made the content or from what source, so you always have to trace the origin yourself.

+What age is this 'Bias, Fairness, Provenance' lesson for?

It is a Stage 6 lesson for roughly ages 11–12 in KAVI Academy's AI-literacy track, the top of primary and secondary-ready. It builds on earlier lessons about trusted sources, confidently-wrong answers, and catching a convincing mistake.

Ready to play and learn?

Start the free interactive lesson, or print the handout for screen-free play.

Start the free lesson →⬇ Download PDF

Part of AI Literacy · Stage 6 · S6 on KAVI Academy — a secular, whole-child learning platform for ages 5–12.