Computer Science · Grade 9 · Networks and the Global Web · Term 2

AI Ethics and Bias

Students will discuss ethical considerations in AI development, including bias, fairness, and accountability.

• Ontario Curriculum Expectations
• CS.HS.IC.2
• CS.HS.S.15

About This Topic

AI ethics and bias addresses the moral challenges in artificial intelligence development, with a focus on how biased data and algorithms produce unfair results. Grade 9 students in Ontario's Computer Science curriculum analyze cases like facial recognition systems that misidentify certain ethnic groups or loan algorithms that disadvantage minorities. This topic fits the Networks and the Global Web unit, as AI powers many online tools and global platforms.

Students tackle key questions, such as how bias enters AI through training data or design choices, its real-world harms like discrimination, ethical duties of developers and users, and creating fairness assessment frameworks. These explorations foster responsible digital citizenship and prepare students for tech's societal roles.

Active learning excels with this topic because ethical issues feel distant until students engage directly. Group debates on AI accountability or hands-on bias audits of sample datasets turn theory into personal insight, encouraging empathy and collaborative problem-solving.

Key Questions

1. Explain how bias can be introduced into AI systems and its potential consequences.
2. Evaluate the ethical responsibilities of AI developers and users.
3. Design a framework for assessing the fairness of an AI-powered decision-making system.

Learning Objectives

• Analyze case studies to identify specific examples of bias in AI systems and explain their origins.
• Evaluate the ethical responsibilities of AI developers and users in mitigating bias and ensuring fairness.
• Design a framework with at least three criteria for assessing the fairness of an AI-powered decision-making system.
• Explain the potential consequences of biased AI on different demographic groups.

Before You Start

Introduction to Artificial Intelligence Concepts

Why: Students need a basic understanding of what AI is and how it learns from data before exploring ethical considerations.

Data Representation and Structures

Why: Understanding how data is organized and processed is foundational to grasping how bias can be embedded within datasets used for AI training.

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.
Fairness in AI	The principle that AI systems should not create or perpetuate unjust discrimination against individuals or groups, ensuring equitable treatment and outcomes.
Accountability in AI	The obligation of AI developers, deployers, and users to take responsibility for the outcomes of AI systems, including addressing errors and harms.
Training Data	The dataset used to train an AI model. Biases present in this data can be learned and amplified by the AI.

Watch Out for These Misconceptions

Common Misconception

AI systems are neutral if trained on large datasets.

What to Teach Instead

Large datasets often amplify societal biases present in real-world data. Group audits of datasets help students spot imbalances firsthand, leading to discussions on diverse data needs. Peer teaching reinforces that size alone does not ensure fairness.

Common Misconception

Bias in AI only comes from intentional developer choices.

What to Teach Instead

Unintentional biases arise from historical data patterns or overlooked assumptions. Role-play activities simulating data collection reveal hidden influences, helping students appreciate systemic issues. Collaborative debriefs build nuanced understanding.

Common Misconception

Ethics discussions are separate from technical computer science skills.

What to Teach Instead

Ethics integrates with coding and design choices. Framework-building tasks show students how fairness metrics fit into algorithms. Hands-on integration makes the connection concrete and relevant to future projects.

Active Learning Ideas

Case Study Rotation: Real-World AI Bias

Prepare four cases: facial recognition errors, biased hiring tools, predictive policing, and credit scoring. Small groups rotate through stations every 10 minutes, noting bias sources, impacts, and fixes on worksheets. End with whole-class share-out.

50 min·Small Groups

Debate Pairs: Developer vs. User Responsibility

Pair students to debate if AI bias fixes responsibility lies more with developers or users. Provide evidence cards on data sourcing and deployment. Pairs present arguments, then vote class-wide on strongest points.

40 min·Pairs

Framework Design: Fairness Checklist

In small groups, students review AI scenarios and co-create a fairness checklist covering data diversity, testing, and transparency. Test the checklist on a sample AI tool description, then refine based on peer feedback.

45 min·Small Groups

Dataset Audit: Individual Bias Hunt

Give students sample datasets from public AI projects. Individually, they identify bias indicators like underrepresentation, score severity, and suggest balanced alternatives. Share findings in a gallery walk.

35 min·Individual

Real-World Connections

• Hiring software used by large corporations can inadvertently discriminate against certain candidates if the AI is trained on historical hiring data that reflects past biases.
• Facial recognition technology used by law enforcement has shown higher error rates for individuals with darker skin tones, leading to potential misidentifications and wrongful accusations.
• Loan application algorithms used by financial institutions might unfairly deny credit to applicants from specific neighborhoods or demographic groups based on biased historical lending data.

Assessment Ideas

Discussion Prompt

Present students with a scenario: An AI system is used to recommend job candidates. One group argues it's efficient, another claims it's biased against women. Ask students to facilitate a debate, identifying potential sources of bias and proposing solutions for fairness.

Quick Check

Provide students with a short description of an AI application (e.g., a content recommendation algorithm). Ask them to write down two potential ethical concerns related to bias and one question they would ask the developers about accountability.

Exit Ticket

Students will write one sentence explaining how bias can enter an AI system and one sentence describing a real-world consequence of biased AI. They will also list one ethical responsibility of an AI user.

Frequently Asked Questions

How does bias enter AI systems?

Bias enters through skewed training data reflecting societal inequalities, like underrepresenting groups in image datasets, or via algorithm designs that prioritize certain patterns. Developers may overlook these during rushed builds. Students learn this by auditing real datasets, seeing consequences like discriminatory outputs in hiring or policing tools.

What are ethical responsibilities for AI developers?

Developers must ensure diverse data, rigorous bias testing, transparency in models, and accountability for harms. Users share responsibility by questioning outputs and advocating changes. Classroom debates clarify these duties, helping students weigh trade-offs between innovation and fairness in global networks.

How can active learning help students grasp AI ethics?

Active methods like case study rotations and bias audits make abstract ethics tangible. Students debate responsibilities in pairs or design fairness frameworks in groups, building empathy through real scenarios. These approaches reveal bias sources collaboratively, far beyond lectures, and develop advocacy skills for responsible tech use.

How to assess fairness in AI decision systems?

Use frameworks checking data diversity, error rates across groups, transparency, and impact audits. Students can prototype checklists testing for equity in outputs. Aligns with Ontario standards by linking ethics to practical CS skills, preparing for ethical deployments in networked environments.