Designing Ethical Algorithms
MIS Quarterly Executive June 2019
14 Pages Posted: 23 Oct 2017 Last revised: 27 Nov 2019
Date Written: June 1, 2019
Abstract
While Big Data has garnered attention for its impact on firms and consumers, the algorithms that make sense of the large data sets are recently in the news – and not only for creating value but for being biased and unfair. Algorithms – including artificial intelligence, machine learning, and neural networks – drive critical decisions such as which patient is seen, who is offered insurance, or what street is patrolled by the police. Such algorithmic decisions, like all decisions, are biased and make mistakes. However the question remains, who is responsible for managing those mistakes? What is the appropriate role of algorithms in critical decisions such as being promoted, receiving medical care, or being offered a loan? I turn to focus on the design of ethical algorithms and the responsibility of developers and users of algorithms to ensure algorithms support good decisions.
I make two ethical arguments. First, while mistakes may be unintentional, ignoring or even fostering mistakes is unethical –mistakes can diminish rights, cause harms, or be unfair. Ethical algorithms should be designed to be governed by identifying, judging, and correcting mistakes. Second, by creating inscrutable, autonomous algorithms, developers may voluntarily take on accountability for the role of the algorithm in the decision including the ability to correct mistakes.
Suggested Citation: Suggested Citation