EVOPEDIA

← all systems · see it in The System

Computer ethics

The part of practical philosophy concerned with how computing professionals should make decisions about professional and social conduct (src:wp:computer-ethics); in Maner's founding definition, the study of ethical problems 'aggravated, transformed or created by computer technology' (src:bynum-sep-2015-computer-ethics). Walked as the field describes itself: the effects computing has on people, the cases and audits that bring them to notice, the policy vacuums they open (Moor), the analysis, the codes, principles and review frameworks, the shared values, and the policies, designs, teaching and laws sent back out. AI ethics, professional (software engineering) ethics and ICT research ethics are booked as its kinds.

Properties published scienceour proposal

Sourced fields, and our own readings.

born at1940s roots (Wiener); named 1976 (Maner); a field by 1985
born precisiondecade
boundaryInside: the scholars and practitioners who do computer ethics, their cases and audits, their analysis (analogies, theories, Moor's method), the professional codes and AI guidelines, the research ethics frameworks (Menlo) and review, the shared values and principles, and the policies, designs and teaching they produce. Outside: the computing technology and the people and societies it affects, the law that may or may not adopt its policies, the companies and states that build and deploy systems, and the older ethical traditions it borrows from.
nested levelA branch of applied (practical) ethics, like medical and business ethics (src:bynum-sep-2015-computer-ethics), its foundations investigated by information ethics (src:wp:computer-ethics); practised by people in professions, universities and review boards.
scalesocial

Measures (17) published science

AI ethics guideline documents (to 2019)84 documents
share released after 201688 %
documents by private companies / governmental agencies19 (22.6%) / 18 (21.4%) documents
principle frequency: transparency / justice and fairness / non-maleficence / responsibility / privacy73 / 68 / 60 / 60 / 47 of 84 documents
issuers in the USA and UK20 (23.8%) and 14 (16.7%) documents
code-exposure test: participants105 developers (5+ years), 63 graduate students; 11 scenarios people
effect of seeing the ACM code on responsesno statistically significant difference for any vignette
commercial gender classification: worst error (darker-skinned women) / best-case group error (lighter-skinned men)up to 34.7 / 0.8 %
Black patients flagged for extra help: algorithm as built / with the disparity remedied17.7 / 46.5 %
Therac-25 patients overdosed / deaths contributed to6 / 3 patients
Moral Machine decisions40 million, 233 countries and territories, ten languages decisions
Therac-25 overdose sizeup to 250 times normal dose
NCCV 1991 attendance400 people, 32 US states, 7 other countries
characteristic clock: ACM code revisions1973, 1992, 2018 (19 and 26 years apart)
computing ethics teaching papers reviewed / mentioning assessment100 / 59% papers
characteristic clock: Belmont standards to federal regulationover a decade years
naming to watershed1976 to 1985 years

Open contentions (6) open contention

Where sources or readings disagree. Both stay on the map.

Partners and parts published science

Built from / builds (5)

Sits inside / holds (2)

Trades with (valence) (13)

expressed by (13)

floor filled by (15)

Kinds (9)

Notes published science

Layers

The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.

computer ethics as its scholars, codes and review boards describe it: computing's effects on people up to the policies and designs that answer them (reader's view)

  1. L1: Computing at work in the world: what computers and their makers do to and for people
  2. L2: Noticing: accident reports, audits, surveys, scholars' and practitioners' observations
  3. L3: Naming the problem: policy vacuums and conceptual muddles
  4. L4: Ethical analysis
  5. L5: Codes, principles and review
  6. L6: Shared values and responsibility
  7. L7: Policies, designs, teaching and law sent back out

the boardroom (shared floors)

  1. L4: Ethical analysis: muddles cleared, analogies tested, theories applied (consequentialism, deontology, virtue, Moor's just consequentialism, Wiener's method, Floridi's information ethics)
  2. L5: Codes, principles and review: professional codes (ACM 1973, 1992, 2018; IEEE/ACM Software Engineering Code 1999), AI guidelines, research ethics frameworks (Menlo 2012) and review boards
  3. L6: Shared values and responsibility: core human values, the converging principles (transparency, justice and fairness, non-maleficence, responsibility, privacy), who answers for harm
  4. L7: Action back on the world: policies formulated and justified, designs changed, people taught, laws proposed

way in: Accidents and failures of computer systems

  1. L1: The accident or harm in the world: Therac-25's overdoses (1985-87), the 737 MAX crashes, the USS Vincennes shoot-down; ICT research that harmed people
  2. L2: Investigation: accident reports and analyses (Leveson and Turner on Therac-25; the 737 MAX investigations)
  3. L3: Causes traced across many hands: software errors, poor testing, overconfidence, bad interfaces, dismissed reports; no single person to blame

way in: Audits of what algorithms decide about people

  1. L1: Automated decisions about people: hiring, loans, sentencing risk scores, health-care management, face analysis
  2. L2: Audits: balanced benchmarks (the Pilot Parliaments Benchmark, 1270 people) and outcome data compared across groups
  3. L3: Disparities measured and traced to a cause: error rates by skin type and gender; cost used as a proxy for illness

way in: Ethical traditions and borrowed frameworks

  1. L1: The received policy cluster: a society's existing laws, rules, practices; ethical theories
  2. L2: Analogies: conceptual bridges to past cases, carrying intuitions across
  3. L3: Borrowed frameworks: medical ethics' four principles (principlism) taken up by AI ethics; the Belmont principles restated for ICT research

way in: Guidelines issued by companies, governments and others

  1. L1: Issuers and their priorities: private companies (19 documents), governmental agencies (18), academic bodies (9) and others
  2. L2: Expert committees and corporate teams draft principles
  3. L3: The published corpus: 84 documents, 88% after 2016
  4. L4: Meta-analyses map the corpus: eleven principles, overlaps and omissions

way in: Law and regulation

  1. L1: Statutes and regulations: the Right to Financial Privacy Act (1978), the Computer Matching and Privacy Protection Act (1988), 45 CFR 46 (the Common Rule)
  2. L2: Legal requirements reach researchers and practitioners
  3. L3: Discordant legal regimes across borders: whose law applies on the Internet

way in: New capabilities computing opens

  1. L1: Logically malleable machines: any activity describable as inputs, outputs and logical operations
  2. L2: New choices for action noticed by practitioners, scholars and the public
  3. L3: A policy vacuum named, often with a conceptual muddle

way in: Public moral preferences

  1. L1: People's moral intuitions about machine decisions, across countries
  2. L2: Collected by an online platform (Moral Machine, January 2016 to July 2020): 40 million decisions, 233 countries and territories
  3. L3: Global preferences, individual variation and three clusters of countries

way in: The practitioner's own moral sense

  1. L1: The individual's own personal ethical code
  2. L2: Moral intuitions transferred across an analogy to the new case
  3. L3: Moral imagination brought to design (value sensitive design)

way in: The scholarly conversation

  1. L1: Scholars from many disciplines: computer scientists, philosophers, policy makers, lawyers, journalists, sociologists, psychologists
  2. L2: Journals, special issues, conferences (Metaphilosophy 1985, NCCV 1991, ETHICOMP 1995)
  3. L3: Debates: the uniqueness debate, the Górniak hypothesis, agent- and patient-based theories

way in: The workplace and its codes

  1. L1: Day-to-day work decisions of computing professionals
  2. L2: Any informal code of ethical conduct in the workplace
  3. L3: Exposure to formal codes of ethics
  4. L4: Codes take effect only when embedded in organisational culture and enforced

way in: What the classroom shows

  1. L1: Students in computing courses, stand-alone or with ethics embedded in technical courses
  2. L2: Assessments, self-report surveys, course evaluations, instructor reflections
  3. L3: Published accounts of what worked (100 papers through 2022)

way out: Codes published to practitioners

  1. L1: The practitioner's work decisions
  2. L2: Read by the practitioner (no measurable change in choices in one test)
  3. L3: Published to members, shown in courses and studies
  4. L5: A code drafted and adopted (ACM 1992: 24 statements of personal responsibility; updated July 2018)

way out: Design changed (value sensitive design, privacy by design)

  1. L1: The artefact built to prevent the harm from the start
  2. L4: Conceptual, empirical and technological investigations, iterated
  3. L7: Values of direct and indirect stakeholders chosen

way out: Policies and law proposals

  1. L1: Conduct governed: what may be done with the technology
  2. L5: Taken up as law in some cases (Right to Financial Privacy Act 1978)
  3. L7: Policy formulated to fill the vacuum, then justified

way out: Published critique and audits

  1. L1: Public, press and regulators informed
  2. L6: Findings framed as harms to shared values

way out: Research ethics review

  1. L1: Research run, changed or stopped
  2. L3: Stakeholder analysis and ethical impact assessment; review boards
  3. L5: A framework: respect for persons, beneficence, justice, respect for law and public interest

way out: Teaching

  1. L1: Students who will build computer artefacts
  2. L3: Courses: stand-alone, or ethics embedded in technical courses
  3. L5: Curricula: Maner's starter kit (1978, published 1980), ABET accreditation, the ACM/IEEE-CS/AAAI curricula

Sources (23) published science

Bynum TW (2015 revision; Summer 2020 archive). Computer and Information Ethics. Stanford Encyclopedia of Philosophy.
Pierce JD, Jones AG, Warren MJ (2006). Penetration Testing Professional Ethics: a conceptual model and taxonomy. Australasian Journal of Information Systems 13(2). doi:10.3127/ajis.v13i2.52
Noorman M (2023 revision). Computing and Moral Responsibility. Stanford Encyclopedia of Philosophy. (behind its links)
Algorithmic bias (encyclopedia) (behind its links)
Mittelstadt B (2019). Principles alone cannot guarantee ethical AI. Nature Machine Intelligence 1:501-507. doi:10.1038/s42256-019-0114-4 (arXiv 1906.06668) (behind its links)
Computer ethics (encyclopedia) (behind its links)
Jobin A, Ienca M, Vayena E (2019). The global landscape of AI ethics guidelines. Nature Machine Intelligence 1:389-399. doi:10.1038/s42256-019-0088-2 (arXiv 1906.11668, preprint read) (behind its links)
Dittrich D, Kenneally E et al. (2012). The Menlo Report: Ethical Principles Guiding Information and Communication Technology Research. US DHS Science and Technology Directorate (behind its links)
NC State University news release (2018-10-08) on McNamara A, Smith J, Murphy-Hill E (2018). Does ACM's code of ethics change ethical decision making in software development? ESEC/FSE 2018. doi:10.1145/3236024.3264833 (paper itself not reached: host certificate error) (behind its links)
Hagendorff T (2020). The Ethics of AI Ethics: An Evaluation of Guidelines. Minds and Machines 30:99-120. doi:10.1007/s11023-020-09517-8 (arXiv 1903.03425) (behind its links)
Buolamwini J, Gebru T (2018). Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification. PMLR 81:1-15 (FAT* 2018) (behind its links)
Obermeyer Z, Powers B, Vogeli C, Mullainathan S (2019). Dissecting racial bias in an algorithm used to manage the health of populations. Science 366:447-453. doi:10.1126/science.aax2342 (behind its links)
Awad E et al. (2018). The Moral Machine experiment. Nature 563:59-64. doi:10.1038/s41586-018-0637-6 (behind its links)
Therac-25 (encyclopedia) (behind its links)
Brown N, Xie B, Sarder E, Fiesler C, Wiese ES (2024). Teaching Ethics in Computing: A Systematic Literature Review of ACM Computer Science Education Publications. ACM Transactions on Computing Education (NSF PAR open copy) (behind its links)
Privacy (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Ethics (encyclopedia article, WME snapshot 2026-11-01) (behind its links)
Lessig L (1999). The Law of the Horse: What Cyberlaw Might Teach. Harvard Law Review 113:501 (author's final draft, Berkman Klein Center). (behind its links)
Value sensitive design (encyclopedia) (behind its links)
Menlo Report (encyclopedia) (behind its links)
Ethics of artificial intelligence (encyclopedia) (behind its links)
Hacker ethic (encyclopedia) (behind its links)
Ten Commandments of Computer Ethics (encyclopedia) (behind its links)