THE CORE AIM OF VOCABULARY MASTERY · AI ETHICS VOCABULARY · VALUES → IMPACT → FAIRNESS → AGENCY → RESPONSIBILITY
AI ethics vocabulary is the language used to discuss the values, impacts and responsibilities involved when artificial intelligence affects people and society. Terms such as fairness, bias, transparency, explainability, accountability, autonomy, human agency, beneficence and harm matter because technically successful systems can still create difficult moral choices.
The core aim of vocabulary mastery for AI ethics vocabulary is value-and-impact clarity. Learners should be able to identify whose interests are affected, what values may conflict, how benefits and harms are distributed and who remains responsible for consequential decisions.
This page is the AI Ethics Vocabulary owner inside the eduKateSG Vocabulary hub. For operational safeguards, use AI Safety Vocabulary. For organisational controls, use AI Governance Vocabulary.
Central proposition: AI ethics vocabulary is mastered when abstract values can be connected to specific people, decisions, trade-offs and responsibilities.
The AI Ethics Vocabulary Router
- Fairness: bias, discrimination, parity, equity.
- Transparency: disclosure, traceability, explainability.
- Agency: autonomy, consent, human control, choice.
- Impact: benefit, harm, externality, distribution.
- Responsibility: accountability, duty, remedy, redress.
- Values: dignity, privacy, justice, beneficence.
Bias and Fairness Are Different
Bias can refer to systematic differences in data, models or decisions. Fairness is a normative judgement about what distributions or decision rules are acceptable. A system can satisfy one mathematical fairness criterion while failing another, so context matters.
Transparency and Explainability
Transparency concerns openness about the system, its use, data, limitations or decision process. Explainability concerns whether people can understand why a particular output or behaviour occurred. They overlap but are not identical.
Human Agency
Human agency means preserving meaningful human ability to understand, choose, contest or override decisions. Merely placing a person somewhere in a workflow does not guarantee meaningful control.
Benefit, Harm and Distribution
Ethical analysis asks not only whether a system creates overall benefit, but who receives that benefit and who bears risk or harm. Vocabulary such as stakeholder, externality, equity and redress helps make distribution visible.
How to Learn AI Ethics Vocabulary
- Choose a concrete AI decision.
- List affected stakeholders.
- Identify competing values.
- Describe potential benefits and harms.
- Ask who has meaningful choice.
- Identify accountability and redress.
- Compare alternative designs or policies.
Common Mistakes
Treating ethics as a checklist
Repair: analyse real value conflicts and consequences.
Using bias and unfairness as synonyms
Repair: distinguish measured systematic patterns from normative judgement.
Assuming explanation automatically creates accountability
Repair: identify who has authority and responsibility to act.
Frequently Asked Questions
What is AI ethics vocabulary?
It is the language used to discuss fairness, bias, transparency, human agency, accountability, benefit and harm in AI systems.
What is human agency?
It is meaningful human capacity to understand, choose, contest or control consequential decisions.
What is redress?
It is a mechanism through which affected people can seek correction, remedy or review.
The AI Ethics Vocabulary Standard
Mastery means connecting values to affected stakeholders, concrete impacts, competing interests and responsible decision-makers.
That is the standard: ethical language precise enough to improve choices rather than decorate them.