Knowledge Organiser: Ethical Issues in Computing
Key Terms
- Ethics: The study of what is morally right and wrong (not just what is legal)
- Privacy: An individual's right to control their own personal information
- Surveillance: Monitoring people's activities, raising questions of freedom vs security
- Automation: Using technology to perform tasks previously done by humans, affecting employment
- Digital divide: The gap between those with access to technology and those without
- Algorithmic bias: When AI systems produce unfair outcomes due to biased training data
Must-Know Facts
- Something can be legal but still unethical (and vice versa)
- Ethical questions require balanced discussion — present both benefits and concerns
- Privacy concerns include: data tracking, facial recognition, data selling to third parties
- Automation can increase efficiency but displaces workers, creating unemployment
- Digital divide affects developing countries and disadvantaged groups disproportionately
- Computer scientists must ask "should we build this?" not just "can we?"
Key Concepts
- PADS-B: Privacy, Automation, Digital divide, Surveillance, Bias in AI
- Facial recognition: useful for security but risks mass surveillance and bias
- AI decision-making raises questions of accountability and fairness
- Exam answers: always give BOTH sides with specific examples
Common Mistakes
- Confusing ethical and legal issues: Legal issues concern what the law says is allowed; ethical issues concern what is morally right — something can be legal but unethical (e.g. collecting vast amounts of personal data legally but unfairly)
- Giving one-sided answers: Exam questions on ethical issues always require a balanced discussion — present both the benefits and the concerns with specific named examples
- Describing automation only as job losses: Automation also creates new technology roles and improves efficiency and safety — examiners expect both the negative displacement of workers AND the positive creation of new jobs
- Confusing algorithmic bias with human bias: Algorithmic bias occurs when AI systems trained on biased or unrepresentative data produce unfair outcomes — it is not the same as a person being biased; the algorithm itself perpetuates the unfairness at scale
- Treating the digital divide as just about internet access: The digital divide also encompasses differences in digital skills, device ownership, and affordability — it affects marginalised groups and developing nations, not just geography