AI Bias & Critical Digital Literacy Workshop
Using AI with judgment instead of blind trust.
A live, facilitated workshop on AI bias and critical digital literacy. It asks participants to slow down, question outputs, and recognize how bias can show up in tools that seem neutral.
Graduate team project · Delivered live to 15 participants · October 21, 2025 · EDET 6015
A live, facilitated session with polling, discussion, and a Kahoot! practice round.
Slides in Canva, delivered on Microsoft Teams, reviewed in Kahoot!.
Presentation slides
The full workshop deck
Page through the slides below without leaving this page.
Made in Canva. If the viewer does not load, use the buttons above.
AI Bias Handout
A take-away reference for the eight bias types
Participants left with a handout that gives, for each of the eight bias types, a definition, an example, and a recommended action. It is designed to be reused after the session.
Adapted from Women Talkin’ ‘Bout AI (Director). (2024, August 29). 10 Ways to Recognize and Mitigate AI Bias: A Beginner’s Guide [Video recording]. youtube.com. Any chatbot names in the examples are illustrative only, not endorsements.
Kahoot! review
A quick, anonymous practice round
After the content, participants played a Kahoot! on their own devices to practice recognizing the bias types and choosing a stronger response. It checks understanding without feeling like a test, and it gives the facilitator a live read on what needs another look.
Try the quiz yourself
The quiz opens in Kahoot! in solo mode, so you can play through it at your own pace.
Opens in a new tab on kahoot.it.
Sample questions
All eight questions from the quiz, with the correct answer marked. Open a question to see its options. If you would rather play it fresh, take the quiz first.
Question 1 How does AI change what “being literate” means today?
- You have to read faster
- You have to question algorithms, not just authors Correct answer
- You don’t need to read: AI reads for you
- You just need better Wi-Fi
Question 2 AI keeps calling every idea “groundbreaking.” What’s that?
- Accurate. All my ideas are groundbreaking.
- Linguistic bias: it favors dramatic adjectives Correct answer
- Temporal bias.
- AI trying to get extra credit.
Question 3 Prompt given to ChatGPT: “Create an image of a doctor and a nurse.” What bias is the output displaying?
- Quantification
- Gender/Racial Correct answer
- Temporal
- Source
Question 4 How could you correct the bias in the last image?
- Keep regenerating until it looks fair by chance
- Accept it: AI just reflects real data
- Be intentional and specify diversity or balance in your prompt Correct answer
- Ask to speak to the AI’s manager
Question 5 You catch AI using Reddit for legal advice. What’s the best move?
- Start adding Esquire after your name in email signatures
- Nothing: Reddit is a reputable source
- Challenge ChatGPT to a lawsuit
- Ask for verified legal sources Correct answer
Question 6 AI says, “Most people celebrate Thanksgiving in November.” Which bias is this?
- Majority bias Correct answer
- Simplification bias
- Temporal bias
- Source bias
Question 7 When you use AI to “save time thinking,” who’s actually doing the thinking?
- You, but faster
- The AI
- The people who trained the AI Correct answer
- Mark Zuckerberg
Question 8 AI is trained on millions of sentences. What does that make it an expert in?
- Truth
- Object reasoning
- Winning arguments online
- Human patterns of language Correct answer
Text only. The quiz itself is played on Kahoot!, where each question also has an image.
The session
From a one-word warm-up to applying what they learned.
The session was built around three goals:
- Identify bias in digital information
- Evaluate credibility using practical strategies
- Apply what they learned through interactive activities
Ground rules shared with participants at the start of the session:
- Mute your microphone unless asked to speak
- Keep your camera on
- Have a device ready for a Kahoot!
- Be ready to participate
- All polls are anonymous
- Warm-upParticipants answered a prompt: when you hear digital literacy, what is the first word that comes to mind?
- FoundationsWhat digital literacy is, and where AI fits into it.
- How AI learnsA plain-language look at machine learning: AI takes in data, analyzes it, finds patterns, and makes predictions.
- How AI learns biasHow human bias, biased data, language, and system design can carry into AI output.
- PracticeA Kahoot! round to practice recognizing the bias types.
- ReflectionParticipants considered how they can use critical digital literacy in daily life, and closed on one habit: pause, think critically, and question what you see online.
Bias types
Eight types of bias, in three families.
The session organizes AI bias into three families, so participants have a simple map for what to look for.
Data bias
Temporal bias, majority bias, and race and gender bias: the data a system learns from shapes what it gets right.
Language bias
Linguistic bias, simplification bias, and quantification bias: how words, summaries, and numbers can tilt an answer.
System bias
Algorithmic bias and source bias: how the design of the system, and where its information comes from, affect the output.