Law Commons^™

Regulating Machine Learning: The Challenge Of Heterogeneity, Cary Coglianese

All Faculty Scholarship

Machine learning, or artificial intelligence, refers to a vast array of different algorithms that are being put to highly varied uses, including in transportation, medicine, social media, marketing, and many other settings. Not only do machine-learning algorithms vary widely across their types and uses, but they are evolving constantly. Even the same algorithm can perform quite differently over time as it is fed new data. Due to the staggering heterogeneity of these algorithms, multiple regulatory agencies will be needed to regulate the use of machine learning, each within their own discrete area of specialization. Even these specialized expert agencies, though, …

Creating Data From Unstructured Text With Context Rule Assisted Machine Learning (Craml), Stephen Meisenbacher, Peter Norlander

School of Business: Faculty Publications and Other Works

Popular approaches to building data from unstructured text come with limitations, such as scalability, interpretability, replicability, and real-world applicability. These can be overcome with Context Rule Assisted Machine Learning (CRAML), a method and no-code suite of software tools that builds structured, labeled datasets which are accurate and reproducible. CRAML enables domain experts to access uncommon constructs within a document corpus in a low-resource, transparent, and flexible manner. CRAML produces document-level datasets for quantitative research and makes qualitative classification schemes scalable over large volumes of text. We demonstrate that the method is useful for bibliographic analysis, transparent analysis of proprietary data, …

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 …

Digital Cluster Markets, Herbert J. Hovenkamp

All Faculty Scholarship

This paper considers the role of “cluster” markets in antitrust litigation, the minimum requirements for recognizing such markets, and the relevance of network effects in identifying them.

One foundational requirement of markets in antitrust cases is that they consist of products that are very close substitutes for one another. Even though markets are nearly always porous, this principle is very robust in antitrust analysis and there are few deviations.

Nevertheless, clustering noncompeting products into a single market for purposes of antitrust analysis can be valuable, provided that its limitations are understood. Clustering contributes to market power when (1) many customers …

Assessing Automated Administration, Cary Coglianese, Alicia Lai

All Faculty Scholarship

To fulfill their responsibilities, governments rely on administrators and employees who, simply because they are human, are prone to individual and group decision-making errors. These errors have at times produced both major tragedies and minor inefficiencies. One potential strategy for overcoming cognitive limitations and group fallibilities is to invest in artificial intelligence (AI) tools that allow for the automation of governmental tasks, thereby reducing reliance on human decision-making. Yet as much as AI tools show promise for improving public administration, automation itself can fail or can generate controversy. Public administrators face the question of when exactly they should use automation. …

The Input Fallacy, Talia B. Gillis

Faculty Scholarship

Algorithmic credit pricing threatens to discriminate against protected groups. Traditionally, fair lending law has addressed such threats by scrutinizing inputs. But input scrutiny has become a fallacy in the world of algorithms.

Using a rich dataset of mortgages, I simulate algorithmic credit pricing and demonstrate that input scrutiny fails to address discrimination concerns and threatens to create an algorithmic myth of colorblindness. The ubiquity of correlations in big data combined with the flexibility and complexity of machine learning means that one cannot rule out the consideration of protected characteristics, such as race, even when one formally excludes them. Moreover, using …

Data Privacy Issues In West Virginia And Beyond: A Comprehensive Overview, Jena Martin

Consumer Law Scholarship

This white paper was commissioned by the Center for Consumer Law and Education, a joint initiative launched by West Virginia University and Marshall University to “coordinate the development of consumer law, policy, and education research to support and serve consumers.”

As such, this paper has a dual purpose. First, it provides a comprehensive overview of the many different legal issues that affect data privacy concerns (both nationally and in West Virginia). Second, it documents and discusses the result of a survey and specific focus groups that were undertaken throughout the fall of 2019 into January 2020 where individuals within the …

Ad Tech & The Future Of Legal Ethics, Seth Katsuya Endo

UF Law Faculty Publications

Privacy scholars have extensively studied online behavioral advertising, which uses Big Data to target individuals based on their characteristics and behaviors. This literature identifies several new risks presented by online behavioral advertising and theorizes about how consumer protection law should respond. A new wave of this scholarship contemplates applying fiduciary duties to information-collecting entities like Facebook and Google.

Meanwhile, lawyers—quintessential fiduciaries—already use online behavioral advertising to find clients. For example, a medical malpractice firm directs its advertising to Facebook users who are near nursing homes with bad reviews. And, in 2020, New York became the first jurisdiction to approve lawyers’ …

The Right To Contest Ai, Margot E. Kaminski, Jennifer M. Urban

Publications

Artificial intelligence (AI) is increasingly used to make important decisions, from university admissions selections to loan determinations to the distribution of COVID-19 vaccines. These uses of AI raise a host of concerns about discrimination, accuracy, fairness, and accountability.

In the United States, recent proposals for regulating AI focus largely on ex ante and systemic governance. This Article argues instead—or really, in addition—for an individual right to contest AI decisions, modeled on due process but adapted for the digital age. The European Union, in fact, recognizes such a right, and a growing number of institutions around the world now call for …

"Slack" In The Data Age, Shu-Yi Oei, Diane M. Ring

Faculty Scholarship

This Article examines how increasingly ubiquitous data and information affect the role of “slack” in the law. Slack is the informal latitude to break the law without sanction. Pockets of slack exist for various reasons, including information imperfections, enforcement resource constraints, deliberate nonenforcement of problematic laws, politics, biases, and luck. Slack is important in allowing flexibility and forbearance in the legal system, but it also risks enabling selective and uneven enforcement. Increasingly available data is now upending slack, causing it to contract and exacerbating the risks of unfair enforcement.

This Article delineates the various contexts in which slack arises and …

Administrative Law In The Automated State, Cary Coglianese

All Faculty Scholarship

In the future, administrative agencies will rely increasingly on digital automation powered by machine learning algorithms. Can U.S. administrative law accommodate such a future? Not only might a highly automated state readily meet longstanding administrative law principles, but the responsible use of machine learning algorithms might perform even better than the status quo in terms of fulfilling administrative law’s core values of expert decision-making and democratic accountability. Algorithmic governance clearly promises more accurate, data-driven decisions. Moreover, due to their mathematical properties, algorithms might well prove to be more faithful agents of democratic institutions. Yet even if an automated state were …

The Law Of Employee Data: Privacy, Property, Governance, Matthew T. Bodie

All Faculty Scholarship

The availability of data related to the employment relationship has ballooned into an unruly mass of personal characteristics, performance metrics, biometric recordings, and creative output. The law governing this collection of information has been awkwardly split between privacy regulations and intellectual property rights, with employees generally losing on both ends. This Article rejects a binary approach that either carves out private spaces ineffectually or renders data into isolated pieces of ownership. Instead, the law should implement a hybrid system that provides workers with continuing input and control without blocking efforts at joint production. In addition, employers should have fiduciary responsibilities …

Towards A Data-Driven Financial System: The Impact Of Covid-19, Nydia Remolina

Centre for AI & Data Governance

The COVID-19 outbreak has a growing impact on the global economy and the financial sector, which plays a critical role in mitigating the unprecedented macroeconomic and financial shock caused by the pandemic. Given the unprecedented nature of the current crisis, financial regulators and supervisors, central banks, along with governments and legislatures face challenges to maintain financial stability, preserve the well-functioning core markets, and ensure the flow of credit to the real economy. Even though the COVID-19 has slowed down our daily lives and stopped the operation of many industries, it did not have the same effect in the data-driven finance …

A New Compact For Sexual Privacy, Danielle K. Citron

Faculty Scholarship

Intimate life is under constant surveillance. Firms track people’s periods, hot flashes, abortions, sexual assaults, sex toy use, sexual fantasies, and nude photos. Individuals hardly appreciate the extent of the monitoring, and even if they did, little can be done to curtail it. What is big business for firms is a big risk for individuals. The handling of intimate data undermines the values that sexual privacy secures—autonomy, dignity, intimacy, and equality. It can imperil people’s job, housing, insurance, and other crucial opportunities. More often, women and minorities shoulder a disproportionate amount of the burden.

Privacy law is failing us. Our …

Ethics, Ai, Mass Data And Pandemic Challenges: Responsible Data Use And Infrastructure Application For Surveillance And Pre-Emptive Tracing Post-Crisis, Mark Findlay, Jia Yuan Loke, Nydia Remolina Leon, Yum Yin, Benjamin (Tan Renyan) Tham

Research Collection Yong Pung How School Of Law

As the COVID-19 health pandemic rages governments and private companies across the globe are utilising AI-assisted surveillance, reporting, mapping and tracing technologies with the intention of slowing the spread of the virus. These technologies have the capacity to amass personal data and share for community control and citizen safety motivations that empower state agencies and inveigle citizen co-operation which could only be imagined outside such times of real and present danger. While not cavilling with the short-term necessity for these technologies and the data they control, process and share in the health regulation mission, this paper argues that this infrastructure …

Mining The Harvard Caselaw Access Project, Felix B. Chang, Erin Mccabe, James Lee

Faculty Articles and Other Publications

This Essay illustrates how machine learning can disrupt legal scholarship through the algorithmic extraction and analysis of big data. Specifically, we utilize data from Harvard Law School’s Caselaw Access Project to model how courts tackle two thorny question in antitrust: the measure of market power and the balance between antitrust and regulation.

Big Data Prosecution And Brady, Andrew Ferguson

Articles in Law Reviews & Other Academic Journals

Prosecutors are joining the big data revolution, adopting “intelligence-driven” strategies to target crime patterns. Centralized big data systems now track offenders, places, and groups allowing prosecutors to link crimes by time, place, associations, or other connections. Adding to these types of formalized, structured databases are growing sources of raw, unstructured big data from digital surveillance technologies like video cameras, police body cameras, and automated license plate readers. The prosecutors of the future will sit on a wealth of valuable investigative insights – all searchable and potentially relevant for a more aggressive and proactive investigation strategy.But as helpful as these new …

Tracking Client Outcomes: A Qualitative Assessment Of Civil Legal Aid’S Use Of Outcomes Data, With Recommendations, David Udell, Amy Widman

Faculty Scholarship

In virtually all sectors of society, people are using data to improve what they do. Everyone, it seems, is interested in data, and is searching for best strategies to draw on its power. The stakes are high in the civil legal aid community, where strengthened advocacy can enable people to preserve their homes, their relationships with their children, their life savings, their physical and emotional well-being, and even their freedom.

Yet, in the civil legal aid community, awareness of the power of data is just beginning to take root. Traditionally, civil legal aid has been thinly funded, with little infrastructure …

Biobanks As Innovation Infrastructure For Translational Medicine, W. Nicholson Price Ii

Book Chapters

Biobanks represent an opportunity for the use of big data to drive translational medicine. Precision medicine demands data to shape treatments to individual patient characteristics; large datasets can also suggest new uses for old drugs or relationships between previously unlinked conditions. But these tasks can be stymied when data are siloed in different datasets, smaller biobanks, or completely proprietary private resources. This hampers not only analysis of the data themselves, but also efforts to translate data-based insights into actionable recommendations and to transfer the discovered technology into a commercialization pipeline. Cross-project technological innovation, development, and validation are all more difficult …

The Ironic Privacy Act, Margaret Hu

Scholarly Articles

This Article contends that the Privacy Act of 1974, a law intended to engender trust in government records, can be implemented in a way that inverts its intent. Specifically, pursuant to the Privacy Act's reporting requirements, in September 2017, the U.S. Department of Homeland Security (DHS) notified the public that record systems would be modified to encompass the collection of social media data. The notification justified the collection of social media data as a part of national security screening and immigration vetting procedures. However, the collection will encompass social media data on both citizens and noncitizens, and was not explicitly …

Hardware, Heartware, Or Nightmare: Smart-City Technology And The Concomitant Erosion Of Privacy, Leila Lawlor

Scholarly Articles

Smart-city technology is being adopted in cities all around the world to simplify our lives, save us time, ease traffic, improve education, reduce energy usage, and keep us healthy and safe. Its adoption is necessary because of changes that are predicted for urban dwellers over the next three decades; urban population and travel are predicted to increase dramatically and our population is graying, meaning the population will include a much greater number of elderly citizens. As these changes occur, smart-city technology can have a huge impact on public safety, improving the ability of law enforcement to investigate crimes, both with …

Toward The Personalization Of Copyright Law, Adi Libson, Gideon Parchomovsky

All Faculty Scholarship

In this Article, we provide a blueprint for personalizing copyright law in order to reduce the deadweight loss that stems from its universal application to all users, including those who would not have paid for it. We demonstrate how big data can help identify inframarginal users, who would not pay for copyrighted content, and we explain how copyright liability and remedies should be modified in such cases.

Fintech And The Innovation Trilemma, Yesha Yadav, Chris Brummer

Vanderbilt Law School Faculty Publications

Whether in response to roboadvising, artificial intelligence, or crypto-currencies like Bitcoin, regulators around the world have made it a top policy priority to supervise the exponential growth of financial technology (or "fintech") in the post-Crisis era. However, applying traditional regulatory strategies to new technological ecosystems has proven conceptually difficult. Part of the challenge lies in the tradeoffs involved in regulating innovations that could conceivably both help and hurt consumers and market participants alike. Problems also arise from the common assumption that today's fintech is a mere continuation of the story of innovation that has shaped finance for centuries.

This Article …

Exploring The Interfaces Between Big Data And Intellectual Property Law, Daniel J. Gervais

Vanderbilt Law School Faculty Publications

This article reviews the application of several IP rights (copyright, patent, sui generis database right, data exclusivity and trade secret) to Big Data. Beyond the protection of software used to collect and process Big Data corpora, copyright’s traditional role is challenged by the relatively unstructured nature of the non-relational (noSQL) databases typical of Big Data corpora. This also impacts the application of the EU sui generis right in databases. Misappropriation (tort-based) or anti-parasitic behaviour protection might apply, where available, to data generated by AI systems that has high but short-lived value. Copyright in material contained in Big Data corpora must …

A Note On Science, Legal Research And Artificial Intelligence, Sean Goltz, Giulia Dondoli

Research outputs 2014 to 2021

This paper discusses the principles of scientific research and in turn review legal research that was done using Artificial Intelligence arguing that it is the tools (Artificial Intelligence) that take center stage while the meaning (legal research) is left back stage. In turn, this kind of research does not adhere to the fundamentals of scientific research nor comply with scientific and industry ethical codes.

Antidiscriminatory Algorithms, Stephanie Bornstein

UF Law Faculty Publications

Can algorithms be used to advance equality goals in the workplace? A handful of legal scholars have raised concerns that the use of big data at work may lead to protected class discrimination that could fall outside the reach of current antidiscrimination law. Existing scholarship suggests that, because algorithms are “facially neutral,” they pose no problem of unequal treatment. As a result, algorithmic discrimination cannot be challenged using a disparate treatment theory of liability under Title VII of the Civil Rights Act of 1964 (Title VII). Instead, it presents a problem of unequal outcomes, subject to challenge using Title VII’s …

Transparency And Algorithmic Governance, Cary Coglianese, David Lehr

All Faculty Scholarship

Machine-learning algorithms are improving and automating important functions in medicine, transportation, and business. Government officials have also started to take notice of the accuracy and speed that such algorithms provide, increasingly relying on them to aid with consequential public-sector functions, including tax administration, regulatory oversight, and benefits administration. Despite machine-learning algorithms’ superior predictive power over conventional analytic tools, algorithmic forecasts are difficult to understand and explain. Machine learning’s “black-box” nature has thus raised concern: Can algorithmic governance be squared with legal principles of governmental transparency? We analyze this question and conclude that machine-learning algorithms’ relative inscrutability does not pose a …

Privacy In The Age Of Medical Big Data, W. Nicholson Price Ii, I. Glenn Cohen

Articles

Big data has become the ubiquitous watch word of medical innovation. The rapid development of machine-learning techniques and artificial intelligence in particular has promised to revolutionize medical practice from the allocation of resources to the diagnosis of complex diseases. But with big data comes big risks and challenges, among them significant questions about patient privacy. Here, we outline the legal and ethical challenges big data brings to patient privacy. We discuss, among other topics, how best to conceive of health privacy; the importance of equity, consent, and patient governance in data collection; discrimination in data uses; and how to handle …

The Promises And Perils Of Using Big Data To Regulate Nonprofits, Lloyd Histoshi Mayer

Journal Articles

For the optimist, government use of “Big Data” involves the careful collection of information from numerous sources. The government then engages in expert analysis of those data to reveal previously undiscovered patterns. Discovering patterns revolutionizes the regulation of criminal behavior, education, health care, and many other areas. For the pessimist, government use of Big Data involves the haphazard seizure of information to generate massive databases. Those databases render privacy an illusion and result in arbitrary and discriminatory computer-generated decisions. The reality is, of course, more complicated. On one hand, government use of Big Data may lead to greater efficiency, effectiveness, …

Automation And Predictive Analytics In Patent Prosecution: Uspto Implications And Policy, Tabrez Y. Ebrahim

Faculty Scholarship

Artificial-intelligence technological advancements bring automation and predictive analytics into patent prosecution. The information asymmetry between inventors and patent examiners is expanded by artificial intelligence, which transforms the inventor-examiner interaction to machine-human interactions. In response to automated patent drafting, automated office-action responses, "cloems" (computer-generated word permutations) for defensive patenting, and machine-learning guidance (based on constantly updated patent-prosecution big data), the United States Patent and Trademark Office (USPTO) should reevaluate patent-examination policy from economic, fairness, time, and transparency perspectives. By conceptualizing the inventor-examiner relationship as a "patenting market," economic principles suggest stronger efficiencies if both inventors and the USPTO have better information …