Law Commons^™

Discredited Data, Ngozi Okidegbe

Faculty Scholarship

Jurisdictions are increasingly employing pretrial algorithms as a solution to the racial and socioeconomic inequities in the bail system. But in practice, pretrial algorithms have reproduced the very inequities they were intended to correct. Scholars have diagnosed this problem as the biased data problem: pretrial algorithms generate racially and socioeconomically biased predictions, because they are constructed and trained with biased data.

This Article contends that biased data is not the sole cause of algorithmic discrimination. Another reason pretrial algorithms produce biased results is that they are exclusively built and trained with data from carceral knowledge sources – the police, pretrial …

Using Artificial Intelligence In The Law Review Submissions Process, Brenda M. Simon

Faculty Scholarship

The use of artificial intelligence to help editors examine law review submissions may provide a way to improve an overburdened system. This Article is the first to explore the promise and pitfalls of using artificial intelligence in the law review submissions process. Technology-assisted review of submissions offers many possible benefits. It can simplify preemption checks, prevent plagiarism, detect failure to comply with formatting requirements, and identify missing citations. These efficiencies may allow editors to address serious flaws in the current selection process, including the use of heuristics that may result in discriminatory outcomes and dependence on lower-ranked journals to conduct …

The Role Of Data For Ai Startup Growth, James Bessen, Stephen Michael Impink, Lydia Reichensperger, Robert Seamans

Faculty Scholarship

Artificial intelligence (“AI”)-enabled products are expected to drive economic growth. Training data are important for firms developing AI-enabled products; without training data, firms cannot develop or refine their algorithms. This is particularly the case for AI startups developing new algorithms and products. However, there is no consensus in the literature on which aspects of training data are most important. Using unique survey data of AI startups, we find that startups with access to proprietary training data are more likely to acquire venture capital funding.

Toward Evidence-Based Antiracist Policymaking: Problems And Proposals For Better Racial Data Collection And Reporting, Neda Khoshkhoo, Aviva Geiger Schwarz, Luisa Godinez Puig, Caitlin Glass, Geoffrey S. Holtzman, Elaine O. Nsoesie, Jasmine Gonzales Rose

Faculty Scholarship

The study of data concerning racial and ethnic inequities and disparities allows us to better understand experiences of racism, and to see more clearly how and where racism manifests. Studying the effects of racism, in turn, allows us to more easily identify racist policies, so that we can craft antiracist interventions.

Existing race and ethnicity data collection efforts are riddled with gaps and errors, including missing and incomplete data, insufficiently disaggregated data, lack of meaningful longitudinal data, infrequently updated data, non-standardized methodologies, and other problems. These deficiencies significantly hinder evidence-based antiracist policymaking.

This policy report examines the state of racial …

Ethical Ai Development: Evidence From Ai Startups, James Bessen, Stephen Michael Impink, Lydia Reichensperger, Robert Seamans

Faculty Scholarship

Artificial Intelligence startups use training data as direct inputs in product development. These firms must balance numerous trade-offs between ethical issues and data access without substantive guidance from regulators or existing judicial precedence. We survey these startups to determine what actions they have taken to address these ethical issues and the consequences of those actions. We find that 58% of these startups have established a set of AI principles. Startups with data-sharing relationships with high-technology firms; that were impacted by privacy regulations; or with prior (non-seed) funding from institutional investors are more likely to establish ethical AI principles. Lastly, startups …

Legislating Data Loyalty, Woodrow Hartzog, Neil Richards

Faculty Scholarship

Lawmakers looking to embolden privacy law have begun to consider imposing duties of loyalty on organizations trusted with people’s data and online experiences. The idea behind loyalty is simple: organizations should not process data or design technologies that conflict with the best interests of trusting parties. But the logistics and implementation of data loyalty need to be developed if the concept is going to be capable of moving privacy law beyond its “notice and consent” roots to confront people’s vulnerabilities in their relationship with powerful data collectors.

In this short Essay, we propose a model for legislating data loyalty. Our …

The Surprising Virtues Of Data Loyalty, Woodrow Hartzog, Neil M. Richards

Faculty Scholarship

Lawmakers in the United States and Europe are seriously considering imposing duties of data loyalty that implement ideas from privacy law scholarship, but critics claim such duties are unnecessary, unworkable, overly individualistic, and indeterminately vague. This paper takes those criticisms seriously, and its analysis of them reveals that duties of data loyalty have surprising virtues. Loyalty, it turns out, can support collective well-being by embracing privacy’s relational turn; it can be a powerful state of mind for reenergizing privacy reform; it prioritizes human values rather than potentially empty formalism; and it offers solutions that are flexible and clear rather than …