lnbressa Author: Jessica Fjeld, Adam Nagy; Publisher: Berkman Klein Center; Publication Year: 2020. The following paper describes and explains 9 principles of ethical data practices: Informed consent of data subjects, security of data, anonymization, transparency, diversity, bias, prominence and communication. The principles listed in this framework cover 4 essential values of ethical data practices: Fairness, benefit, openness and reliability…
Author: N/A; Publisher: The University of Queensland; Publication Year: N/A. The following article is written for the University of Queensland in Australia, but the ethical principles outlined in the document are best practice for anyone who collects, maintains, and utilizes data. The first principle states that data usage must be defined, and a cost-benefit analysis must be done for individuals affected by the data. Likewise…
Author: Will Knight; Publisher: Wired; Publication Year: 2019. The following article touches on the idea of bias “proxies,” like address information as a proxy for race. It talks about how the Apple Card was offering lower credit limits for women compared to men even though gender was specifically excluded as a variable. The author argues that we should be including those specific variables so that the…
Author: Will Horvath; Publisher: Data Science Ethics; Publication Year: 2019. The following article discusses how with countries fighting against the growing threat of privacy invasion (and many losing), Europe implemented in 2018 the General Data Protection Regulation (GDPR) that provides the ‘most sweeping protection of user privacy rights yet passed in the world.’ France’s CNIL used the GDPR to fine Google for $57 million…
Author: Will Douglas Heaven; Publisher: MIT Technology Review; Publication Year: 2021. The following article discusses how it is a known fact that biased algorithms affect automatic decision-making processes. To fix this problem, many researchers and start-ups are working to build fairer algorithms. The author, however, claims that building a fair algorithm is not enough because low-income and minority groups represent a very small…
Author: Mark D. Wilkinson et al.; Publisher: Nature; Publication Year: 2016. The following article focuses on the need to improve data handling in order to ensure the reproducibility of results published in scientific literature. It sets forth the FAIR data principles (Findability, Accessibility, Interoperability, and Reproducibility). The article argues that FAIRness is a prerequisite for proper data management and stewardship. The…
Author: Jess Whittlestone et al.; Publisher: Conference on AI, Ethics, and Society; Publication Year: 2019. The following research article argues that there are tensions between common principles guiding ethical decision-making about artificial intelligence (AI). They explain 4 key tensions: service quality versus privacy, accuracy versus fair treatment, personalization versus solidarity, and convenience versus dignity. By analogy with bioethics, the authors also argue…
Author: N/A; Publisher: Western Governors University; Publication Year: 2021. The following article discusses how companies with higher ethical standards have been shown to perform better than companies who fail to consider the ethics of their practice. “73% of professionals say they take an organization’s values into account.” It is important for employees in deciding where to work if the company has values that align with…
Author: Wenhong Chen, Anabel Quan-Haase; Publisher: Social Science Computer Review; Publication Year: 2018. The following article addresses new issues created by big data, such as biases, subjectivities, and forms of oppression. They define 4 major aspects of big data ethics and politics: 1). Potential biases in big data collection and interpretation, 2). Community and citizen concerns of big data (mis)use in public life and for journalistic purposes, 3). Media…
Author: Wendy Hui Kyong Chun; Publisher: MIT Press; Publication Year: 2021. The following book discusses the innate bias that comes from dirty data and how removing indicators of race will not remove bias from artificial intelligence. The author explains several concepts including: alternative algorithms, defaults, and interdisciplinary coalitions to foster more ethical/democratic data…