Solon Barocas

Big data's end run around procedural privacy protections

Recognizing the inherent limitations of consent and anonymity.

Ten simple rules for responsible big data research

The use of big data research methods has grown tremendously over the past five years in both academia and industry. As the size and complexity of available datasets has grown, so too have the ethical questions raised by big data research. These questions become increasingly urgent as data and research agendas move well beyond those typical of the computational and natural sciences, to more directly address sensitive aspects of human behavior, interaction, and health. The tools of big data research are increasingly woven into our daily lives, including mining digital medical records for scientific and economic insights, mapping relationships via social media, capturing individuals’ speech and action via sensors, tracking movement across space, shaping police and security policy via “predictive policing,” and much more.

Engaging the ethics of data science in practice

Seeking more common ground between data scientists and their critics.

Designing Against Discrimination in Online Markets

Platforms that connect users to one another have flourished online in domains as diverse as transportation, employment, dating, and housing. When users interact on these platforms, their behavior may be influenced by preexisting biases, including tendencies to discriminate along the lines of race, gender, and other protected characteristics. In aggregate, such user behavior may result in systematic inequities in the treatment of different groups. While there is uncertainty about whether platforms bear legal liability for the discriminatory conduct of their users, platforms necessarily exercise a great deal of control over how users’ encounters are structured—including who is matched with whom for various forms of exchange, what information users have about one another during their interactions, and how indicators of reliability and reputation are made salient, among many other features. Platforms cannot divest themselves of this power; even choices made without explicit regard for discrimination can affect how vulnerable users are to bias. This Article analyzes ten categories of design and policy choices through which platforms may make themselves more or less conducive to discrimination by users. In so doing, it offers a comprehensive account of the complex ways platforms’ design choices might perpetuate, exacerbate, or alleviate discrimination in the contemporary economy.

WSDM 2016 Workshop on the Ethics of Online Experimentation

Online experimentation is now a core and near-constant part of the operation of a production online service, such as a web search engine or social media service. These are large-scale experiments that involve research subjects often numbering in the hundreds of thousands and wide-ranging, computer-automated variations in experimental treatment. In some cases, the results of online experiments may be of use internally to optimize system performance (for example, a test may be conducted to help make web page layout decisions). In other cases, the results may be of academic interest (for example, an experiment may be conducted to test a hypothesis about human behavior). Because of their rapid deployment and broad impact, online experimentation systems provide an extremely valuable tool for scientists and engineers. Despite this statistical power, in some situations, an online experiment can raise difficult ethical questions. One only needs to revisit the conversations resulting from the Facebook emotional contagion experiment to understand that some experiments may, at the very least, warrant careful review before being conducted. Since this episode, scholarship published mainly in the qualitative research and information law communities indicates that this may not be an isolated incident. Ethical and legal problems probably arise in other online experiments, published or not. As experimentation platforms and users become easily accessible, scientists and practitioners may increasingly put the well-being and trust of end users at risk. In light of these concerns, organizations often review online experiments before they are actually conducted. In production settings, the review process might vary with respect to formality or standards across companies and even groups within companies. When intended or used for academic publication, experiments or data may have undergone inconsistent review processes, some implementing academic-style institutional review boards and others none at all. Although there is a suggestion that service providers are concerned about the wellbeing of end users, the community does not