Computing, Society, and Ethics

Constructing an ethical argument

An ethical argument connects factual claims, values, and a conclusion about action. Separate what a system can do from whether it should be built or deployed. Identify affected people, their interests, the distribution of benefits and burdens, and what remains uncertain. A disagreement can arise from different evidence, different values, or different assumptions about responsibility. Make the connection explicit instead of presenting a personal reaction as a complete justification. Technical feasibility does not settle the normative question.

References: ACM Code of Ethics and Professional Conduct

Three ethical frameworks

Consequentialist reasoning evaluates outcomes, requiring attention to whose welfare counts, uncertainty, and how different harms are compared. Deontological reasoning examines duties, rights, and constraints on action, even when violating them appears beneficial. Virtue ethics asks what character and professional judgment a practice expresses and cultivates. Apply each framework carefully rather than reducing it to a slogan. The frameworks can expose different dimensions of a problem without guaranteeing a single uncontested answer.

References: Stanford Encyclopedia of Philosophy — Virtue Ethics · Stanford Encyclopedia of Philosophy — Deontological Ethics · Stanford Encyclopedia of Philosophy — Consequentialism

Professional obligations and institutional responsibility

Professional codes make public welfare, competence, honesty, privacy, and responsibility part of computing practice. A code is a structured reference for deliberation, not an algorithm that resolves every conflict. Responsibility can be distributed across designers, managers, operators, purchasers, and institutions, but distribution should not become an excuse for inaction. Document a concern with concrete evidence, affected stakeholders, and feasible alternatives. Evaluate what a professional can influence and which organizational incentives sustain the problem.

References: ACM Code of Ethics and Professional Conduct

Worked campus-attendance scenario

Imagine a university proposing face recognition to automate lecture attendance. Possible benefits include less administrative work; burdens include misidentification, biometric data exposure, surveillance pressure, and unequal error rates. A consequentialist analysis compares these effects with less intrusive methods. A rights-based analysis asks about meaningful consent, proportionality, and the ability to contest an error. A virtue-based analysis asks whether the practice reflects care and trustworthy stewardship. A useful conclusion specifies safeguards or an alternative, plus the evidence needed to justify the choice.

References: Stanford Encyclopedia of Philosophy — Virtue Ethics · Stanford Encyclopedia of Philosophy — Deontological Ethics · Stanford Encyclopedia of Philosophy — Consequentialism

Bias, privacy, and knowledge

Data reflects choices about what to observe, label, retain, and exclude. A fairness metric can make one disparity visible while hiding another; select it in light of the decision and affected population. Privacy includes power over collection, inference, use, and sharing, rather than only secrecy. Aggregation and prediction can reveal sensitive facts without directly collecting them. Ask who can challenge a record or decision, whether the system gives an intelligible explanation, and how errors propagate through downstream institutions.

References: ACM Code of Ethics and Professional Conduct

Generative systems and reasoned recommendations

For a generative system, examine provenance, consent, labor, reliability, access, and incentives. A plausible output is not evidence of truth, and automated scale can magnify a small per-instance error. Assign accountability for checking consequential outputs and for correcting harms. Fictional scenarios can reveal values and assumptions, but empirical claims about real systems need real evidence. End an analysis with an actionable recommendation, its tradeoffs, who is responsible for it, and the conditions under which the recommendation should be reconsidered.

References: ACM Code of Ethics and Professional Conduct

Practice & worked answers

Original exercises for this guide.

Public notes & references

Course sources establish the scope above. Reference materials support the explanations; linked lecture notes and books remain with their authors.

• Carleton COMP 4701 — Fall 2026 outline ↗Course scope · Fall 2026
• ACM Code of Ethics and Professional Conduct ↗Professional code
• Carleton COMP 4701 — Fall 2026 ethical-framework and technology topic map ↗Topic map
• Stanford Encyclopedia of Philosophy — Virtue Ethics ↗Authored scholarly article
• Stanford Encyclopedia of Philosophy — Deontological Ethics ↗Authored scholarly article
• Stanford Encyclopedia of Philosophy — Consequentialism ↗Authored scholarly article