Ethical algorithm design
Abstract
In this letter, we summarize the research agenda that we survey in our recent book The Ethical Algorithm , which is intended for a general, nontechnical audience. At a high level, this research agenda proposes formalizing the ethical and social values that we want our algorithms to maintain --- values including privacy, fairness, and explainability --- and then to embed these social values directly into our algorithms as part of their design. This broad research area is most mature in the area of privacy, specifically differential privacy . It is off to a good start in emerging areas like algorithmic fairness, and seems promising for more nebulous goals like explainability, if only we can find the right definitions. Most work in this area to date analyzes algorithms as isolated components, but game-theoretic and economic analysis will become increasingly important as we try and study the effects of algorithmic interventions in larger sociotechnical systems.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Some in HCI research embrace new creative tools to create AI Art, but others are skeptical of the hype from different perspectives, including disturbance of social orders, legality of style appropriation, and potential copyright infringement.
“Specifically in the field of HCI research, some embrace these new creative tools to create various AI Art, but others are skeptical of the hype from different perspectives [17], including the disturbance of social orders [44], the legality of style appropriation [38], and potential copyright infringement [18].”
Cites
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137 · cited 4× · see citing works
“2018], or medical costs as a proxy for severity of disease in healthcare applications [Obermeyer et al. 2019].”
BackgroundMethodInfluentialFair prediction with disparate impact: a study of bias in recidivism prediction instruments
Chouldechova, A. (2017) · Big Data · doi:10.1089/big.2016.0047 · Cited by 2,159 · see citing works
Fairness in Criminal Justice Risk Assessments: The State of the Art
Berk, R., Heidari, H., Jabbari, S., Kearns, M., & Roth, A. (2021) · Sociological Methods & Research · doi:10.1177/0049124118782533 · Cited by 988 · cited 1× · see citing works
“…often involve proxies for what we actually want to predict, instead of the real underlying values: arrests as a proxy for crime in criminal risk prediction applications [Berk et al. 2018], or medical costs as a proxy for severity of disease in healthcare applications [Obermeyer et al. 2019].”
BackgroundSelf-citationFairness in Learning: Classic and Contextual Bandits
Joseph, M., Kearns, M., Morgenstern, J., & Roth, A. (2016) · arXiv · doi:10.48550/arxiv.1605.07139 · Cited by 199 · cited 1× · see citing works
“…is also a divide between “statistical” notions of fairness that are already practical but provide limited promises to individuals, and definitions that provide stronger individual fairness guarantees [Dwork et al. 2012; Joseph et al. 2016] but which have various barriers to realistic implementation.”
BackgroundSelf-citation
Flagged first, then by citations
Cited by
Sorted by citations
Dedema, M., & Rosenbaum, H. (2024) · Journal of the Association for Information Science and Technology · doi:10.1002/asi.24868 · Cited by 49 · see citing works
Hagendorff, T. (2021) · Minds and Machines · doi:10.1007/s11023-021-09573-8 · Cited by 47 · see citing works
He, Q., Zheng, W., Bao, H., Chen, R., & Tong, X. (2023) · Proceedings of the Eleventh International Symposium of Chinese CHI · doi:10.1145/3629606.3629675 · Cited by 15 · see citing works
- Authors
- Kearns, M., & Roth, A.
- Affiliations
- University of Pennsylvania (US)
- Venue
- ACM SIGecom Exchanges
- Published
- 2020-12-02
- Volume
- 18
- Issue
- 1
- Pages
- 31-36
- ISSN
- 1551-9031 (ISSN-L)
- OpenAlex
- W3109041932
- Cited by
- 21
- References
- 4
- Citations by year
- 2026: 1 · 2025: 2 · 2024: 4 · 2023: 5 · 2022: 4 · 2021: 4 · 2020: 1
- FWCI
- 4.77
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- ACM SIGecom Exchanges (publishedVersion)
- Topics
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Keywords
- ethical algorithms, algorithm design, privacy, differential privacy, algorithmic fairness, explainability, game-theoretic analysis, economic analysis, socio-technical systems
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
