Bias in AI and Algorithmic Fairness

Active learning works for this topic because bias in AI is not a theoretical issue. It shows up in real datasets and tools students use daily. When students examine concrete examples and redesign scenarios, they see how data choices affect people’s lives in ways that passive lectures cannot match.

• Secondary 3
• Computing
• 4 activities
• 25 min–50 min

Learning Objectives

1. Analyze how specific biases in training data, such as demographic underrepresentation, can lead to discriminatory outcomes in AI applications like facial recognition systems.
2. Evaluate the effectiveness of different strategies, such as data augmentation and algorithmic debiasing, in mitigating AI bias.
3. Design a hypothetical AI system, detailing its purpose, potential biases, and proposed fairness interventions.
4. Justify the necessity of ongoing AI system audits and transparency mechanisms for ensuring equitable societal impact.

Socratic Seminar

⏱ 45 min·Small Groups

Small Groups: Real-World Case Audit

Provide groups with a case study on biased AI, like facial recognition failures. Students identify bias sources in data, evaluate impacts, and propose three fairness fixes. Groups share audits in a class gallery walk.

Prepare & details

Analyze how biases in training data can lead to discriminatory AI outcomes.

Facilitation Tip: During the Real-World Case Audit, assign each small group a different dataset to analyze so the class covers multiple types of bias in one lesson.

Setup: Chairs arranged in two concentric circles

Materials: Discussion question/prompt (projected), Observation rubric for outer circle

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Socratic Seminar

⏱ 35 min·Pairs

Pairs: Bias Scenario Redesign

Pairs design a hypothetical AI system with embedded bias leading to social injustice, then redesign it for fairness using strategies like balanced datasets. They sketch system flows and present changes.

Prepare & details

Justify the importance of auditing AI systems for fairness and transparency.

Facilitation Tip: For the Bias Scenario Redesign, provide sentence starters to help pairs articulate which features of the scenario reveal bias and how their redesign addresses it.

Setup: Chairs arranged in two concentric circles

Materials: Discussion question/prompt (projected), Observation rubric for outer circle

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Socratic Seminar

⏱ 50 min·Whole Class

Whole Class: Fairness Debate

Divide class into teams to debate mandatory AI audits: one side argues benefits for equity, the other potential innovation barriers. Use structured turns and vote on strongest points.

Prepare & details

Design a hypothetical scenario where AI bias could lead to significant social injustice.

Facilitation Tip: In the Fairness Debate, assign roles clearly and give students two minutes to prepare opening statements using facts from their previous activities.

Setup: Chairs arranged in two concentric circles

Materials: Discussion question/prompt (projected), Observation rubric for outer circle

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Socratic Seminar

⏱ 25 min·Individual

Individual: Personal Bias Checklist

Students create a checklist for auditing AI fairness, drawing from class learnings. Test it on a provided algorithm example and reflect on one improvement.

Prepare & details

Analyze how biases in training data can lead to discriminatory AI outcomes.

Facilitation Tip: During the Personal Bias Checklist, model how to reflect on one item from the checklist before asking students to complete the rest independently.

Setup: Chairs arranged in two concentric circles

Materials: Discussion question/prompt (projected), Observation rubric for outer circle

• Analyze
• Evaluate
• Create
• Social Awareness
• Relationship Skills

Teaching This Topic

Experienced teachers approach this topic by grounding abstract concepts in local examples students recognize. They avoid overwhelming students with technical jargon, instead focusing on how data choices impact real people. Research shows that structured discussions after hands-on tasks build deeper understanding than lectures alone.

What to Expect

Successful learning looks like students explaining how biased data leads to unfair outcomes, proposing concrete fairness fixes, and justifying their choices with evidence from the case studies. They should move from noticing problems to suggesting actionable solutions.

Watch Out for These Misconceptions

Common Misconception

During the Real-World Case Audit, watch for students assuming AI is unbiased because it uses math.

What to Teach Instead

Use the dataset handouts to guide students to identify human sources of bias in the data, such as underrepresentation or mislabeling, and discuss why these affect algorithmic outcomes.

Common Misconception

During the Bias Scenario Redesign, watch for students believing better coding can remove all bias.

What to Teach Instead

Have pairs list trade-offs of their redesigns, such as reduced accuracy for certain groups, to highlight that fairness often requires ongoing adjustments rather than perfect solutions.

Common Misconception

During the Fairness Debate, watch for students dismissing AI bias as irrelevant to Singapore.

What to Teach Instead

Use Singapore-based case studies in the debate prompts and ask students to find local examples of biased AI, ensuring connections to their own context.

Assessment Ideas

Discussion Prompt

After the Real-World Case Audit, present the loan application scenario. Ask students to identify biases in the training data, explain how these biases could lead to unfair rejections, and suggest fairness steps the bank should take, using evidence from their case audits.

Exit Ticket

During the Personal Bias Checklist, ask students to write one potential source of bias in the AI tutor example and one strategy to make it fairer, collecting responses as they leave to check understanding of data sources and mitigation strategies.

Quick Check

After the Fairness Debate, display the list of AI fairness strategies (e.g., data diversification, bias detection tools, human oversight) and ask students to match each strategy to a brief description of how it helps mitigate bias, using a short quiz or drag-and-drop activity.

Extensions & Scaffolding

• Challenge: Ask students who finish early to research Singapore’s AI governance guidelines and compare them with fairness strategies they designed.
• Scaffolding: For students struggling with scenario redesign, provide a partially completed template with two bias sources filled in as examples.
• Deeper exploration: Invite students to interview a local professional using AI tools and present findings on bias mitigation in practice.

Key Vocabulary

Algorithmic Bias	Systematic and repeatable errors in a computer system that create unfair outcomes, such as privileging one arbitrary group of users over others.
Training Data	The dataset used to train an AI model. Biases present in this data can be learned and perpetuated by the model.
Fairness Metrics	Quantitative measures used to assess whether an AI system's outcomes are equitable across different demographic groups.
Algorithmic Auditing	The process of examining an AI system's algorithms and data to identify and address potential biases and ensure fairness.
Transparency	The principle of making AI systems' decision-making processes understandable and accessible, allowing for scrutiny and accountability.