The ethics of algorithms: Mapping the debate
By Brent Daniel Mittelstadt, Patrick Allo, Mariarosaria Taddeo, Sandra Wachter, Luciano Floridi. 2016
2,485 citations recorded by the source corpus.
In information societies, operations, decisions and choices previously left to humans are increasingly delegated to algorithms, which may advise, if not decide, about how data should be interpreted and what actions should be taken as a result. More and more often, algorithms mediate social processes, business transactions, governmental decisions, and how we perceive, understand, and interact among ourselves and with the environment. Gaps between the design and operation of algorithms and our understanding of their ethical implications can have severe consequences affecting individuals as well as groups and whole societies. This paper makes three contributions to clarify the ethical importance of algorithmic mediation. It provides a prescriptive map to organise the debate. It reviews the current discussion of ethical aspects of algorithms. And it assesses the available literature in order to identify areas requiring further work to develop the ethics of algorithms.
Subfields: Health Informatics, Public Health, Environmental and Occupational Health, Safety Research, AI Ethics, Fairness & Society
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Health Informatics
- Emilio Ferrara
- Daniel S. Schiff
- Timo Speith
- Kirsten E. Martin
- Luciano Floridi
- Maranke Wieringa
- Marcel A. J. van Gerven
- Gabrielle Ras
In Public Health, Environmental and Occupational Health
- Emilio Ferrara
- Kirsten E. Martin
- Brent Daniel Mittelstadt
- Mariarosaria Taddeo
- Luciano Floridi
- Sandra Wachter
- Fernando Diaz
- Carlos Castillo
In Safety Research
- Daniel S. Schiff
- Luciano Floridi
- Emilio Ferrara
- Jessica Morley
- Ben Hutchinson
- Timo Speith
- Kirsten E. Martin
- Josh Cowls
In AI Ethics, Fairness & Society
- Emilio Ferrara
- Kirsten E. Martin
- Luciano Floridi
- David Watson
- Alexandra Olteanu
- Brent Daniel Mittelstadt
- Simon Schafheitle
- Jan C. Weyerer
Citations in the corpus
- Towards Accountability for Machine Learning Datasets: Practices from Software Engineering and Infrastructure (2020, background): “ how consent is obtained, and the data collection process needs to keep “the receipts” as evidence of due process. To facilitate this, it is therefore necessary that dataset requirements be traceable [100], and in many cases requirements should be measurable. Non-boolean test outcomes are still usef”
- Explanation Methods in Deep Learning: Users, Values, Concerns and Challenges (2018, background): “The following concerns and question are taken from [37] and are adapted to fit the context of the DNN process.”
- Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019, background): “24 Given that our treatment of the subject is purposefully schematic, the interested reader can find more information in related works by Grimmelmann (2015),Metcalf and Crawford (2016), Bowser and Tsai (2015), Benton et al. (2017), and Mittelstadt et al. (2016).”
- Artificial Intelligence and the Implementation Challenge (2019, background): “In cases where training data are partial or incomplete or only reflect a subset of a given population, the resulting model will only be relevant to the population of people represented in the dataset [47].”
- Digital Twins: A Survey on Enabling Technologies, Challenges, Trends and Future Prospects (2022, background): “…inference are probable and therefore an epistemic limitation; (2) traceability between the input data and conclusion is not accessible and open to critique; (2) conclusions are dependent upon the quality of data; or (4) the actions based on conclusions are discriminatory even if well-founded [191].”
- Beyond the individual: governing AI's societal harm (2021, neutral): “These features include, amongst others, AI systems’ self-learning ability and hence potential unpredictability, the vast scale and speed at which they can operate, their ability to distil information from data that may escape the human eye, the opportunity they provide to delegate human authority an”
- Six Human-Centered Artificial Intelligence Grand Challenges (2023, neutral): “Thus, explainable AI has – for good reason – become a prominent keyword in technology ethics (Mittelstadt et al., 2016; Wachter et al., 2017).”
- AI Ethics - Too Principled to Fail? (2019, neutral): “34 This is worrying, as distance from potential victims has been shown to have a positive effect on unethical professional behaviour.”
- Ethical Implications and Accountability of Algorithms (2018, methodology): “Similarly, and focused on algorithms, Mittelstadt et al. note that algorithms are value-laden with biases that are “specified by developers and configured by users with desired outcomes in mind that privilege some values and interests over others”(2016).”
- Education for AI, not AI for Education: The Role of Education and Ethics in National AI Policy Strategies (2021, background): “…and the public have become increasingly aware of issues related to labor displacement (Acemoglu and Restrepo 2019), autonomous vehicles (Bagloee et al. 2016), algorithmic bias (Caliskan et al. 2017), and privacy (Manheim and Kaplan 2018), amongst other issues (Mittelstadt et al. 2016).”