Big Data and Due Process: Toward a Framework to Redress Predictive Privacy Harms

By Kate Crawford, J. Schultz. 2013

470 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Iterative Orthogonal Feature Projection for Diagnosing Bias in Black-Box Models (2016, background): “One key concern is unintentional discrimination (Crawford & Schultz, 2014; Barocas & Selbst, 2014).”
Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019, background): “The reliance on automated decision making processes based on statistical methods, can inherit, propagate, or even amplify the biases and prejudice present in the training data with respect to various factors such as race, age, gender or socioeconomic groups (Crawford and Schultz, 2014; Barocas and S”
Actionable Recourse in Linear Classification (2018, background): “oyment [2] and public services [8, 28]. However, recourse should also be considered more broadly given that a lack of human agency is perceived as a source of injustice in algorithmic decision-making [5,10,21,33]. The potential lack of recourse in algorithmic decision-making is often used to motivat”
Algorithmic surveillance: the collection conundrum (2017, background): “In the general context of algorithmic determinations, Crawford and Schulz argue that the auditing of ‘ the data that was used to make a determination ’ would provide a bene ﬁ cial safeguard (Crawford and Schultz 2014, 117; De Zwart, Humphreys, and Van Dissel 2014, 721).”
Designing a Moral Compass for the Future of Computer Vision Using Speculative Analysis (2017, unknown): “Technology ethics researchers Kate Crawford and Jason Schultz have termed the class of privacy violations that come from unanticipated inference as predictive privacy harms [22].”
Social Injustice in Surveillance Capitalism (2017, background): “As MayerSchönberger and Cukier describe (2013), Facebook’s market valuation at the time of their IPO in 2012 was $104 billion, yet they claimed assets of just $6.3 billion.”
Algorithms at Work: The New Contested Terrain of Control (2020, background): “…encroach on procedural due process, that is, “ the constitutional requirement that any government deprivation of a liberty or property right must be preceded — at a minimum — by notice and the opportunity for a hearing on the matter before an impartial adjudicator ” (Crawford & Schultz, 2014: 111).”
Enslaving the Algorithm: From a “Right to an Explanation” to a “Right to Better Decisions”? (2018, background): “Taken together, these provisions offer exciting opportunities to operationalise what in the US have been called “big data due process” rights [3, 4].”
What the Near Future of Artificial Intelligence Could Be (2019, background): “…or confidentiality at the development stage (though problems persist at the deployment stage, because of the predictive privacy harms (Crawford and Schultz 2014)), they are not immediately sensitive (sensitivity during the deployment stage still matters), if they are lost, it is not a disaster…”
Privacy is an essentially contested concept: a multi-dimensional analytic for mapping privacy (2016, background): “Scholars have argued for privacy concepts and approaches that unburden the individual by attending to the contextual norms of spheres of social life [9,10], address risks posed by ancillary data [11] and attend to the emerging semantics of machine learning [4].”
Big data, bigger dilemmas: A critical review (2015, background): “As leading commentators have observed, “[t]hese types of harms do not necessarily fall within the conventional invasion of privacy boundaries,” and thus raise new questions for policymakers (Crawford & Schultz, 2014, p. 93).”
Where are human subjects in Big Data research? The emerging ethics divide (2016, methodology): “These experiments often use data that may not appear to be personally identifying, but can become so in combination, generating ‘‘predictive privacy harms’’ (Crawford and Schultz, 2014).”
The ethics of smart cities and urban science (2016, background): “Predictive modelling using urban big data can generate inferences about an individual that are not directly encoded in a database but constitute what many would consider to be PII and which produce ‘predictive privacy harms’ [6,51].”