Law Commons^™

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, …

Open-Source Clinical Machine Learning Models: Critical Appraisal Of Feasibility, Advantages, And Challenges, Keerthi B. Harish, W. Nicholson Price Ii, Yindalon Aphinyanaphongs

Articles

Machine learning applications promise to augment clinical capabilities and at least 64 models have already been approved by the US Food and Drug Administration. These tools are developed, shared, and used in an environment in which regulations and market forces remain immature. An important consideration when evaluating this environment is the introduction of open-source solutions in which innovations are freely shared; such solutions have long been a facet of digital culture. We discuss the feasibility and implications of open-source machine learning in a health care infrastructure built upon proprietary information. The decreased cost of development as compared to drugs and …

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 …

Resisting Face Surveillance With Copyright Law, Amanda Levendowski

Georgetown Law Faculty Publications and Other Works

Face surveillance is animated by deep-rooted demographic and deployment biases that endanger marginalized communities and threaten the privacy of all. But current approaches have not prevented its adoption by law enforcement. Some companies have offered voluntary moratoria on selling the technology, leaving many others to fill in the gaps. Legislators have enacted regulatory oversight at the state and city levels, but a federal ban remains elusive. Both approaches require vast shifts in practical and political will, each with drawbacks. While we wait, face surveillance persists. This Article suggests a new possibility: face surveillance is fueled by unauthorized copies and reproductions …

Moving Toward Personalized Law, Cary Coglianese

All Faculty Scholarship

Rules operate as a tool of governance by making generalizations, thereby cutting down on government officials’ need to make individual determinations. But because they are generalizations, rules can result in inefficient or perverse outcomes due to their over- and under-inclusiveness. With the aid of advances in machine-learning algorithms, however, it is becoming increasingly possible to imagine governments shifting away from a predominant reliance on general rules and instead moving toward increased reliance on precise individual determinations—or on “personalized law,” to use the term Omri Ben-Shahar and Ariel Porat use in the title of their 2021 book. Among the various technological, …

Part I - Ai And Data As Medical Devices, W. Nicholson Price Ii

Other Publications

It may seem counterintuitive to open a book on medical devices with chapters on software and data, but these are the frontiers of new medical device regulation and law. Physical devices are still crucial to medicine, but they – and medical practice as a whole – are embedded in and permeated by networks of software and caches of data. Those software systems are often mindbogglingly complex and largely inscrutable, involving artificial intelligence and machine learning. Ensuring that such software works effectively and safely remains a substantial challenge for regulators and policymakers. Each of the three chapters in this part examines …

Algorithm Vs. Algorithm, Cary Coglianese, Alicia Lai

All Faculty Scholarship

Critics raise alarm bells about governmental use of digital algorithms, charging that they are too complex, inscrutable, and prone to bias. A realistic assessment of digital algorithms, though, must acknowledge that government is already driven by algorithms of arguably greater complexity and potential for abuse: the algorithms implicit in human decision-making. The human brain operates algorithmically through complex neural networks. And when humans make collective decisions, they operate via algorithms too—those reflected in legislative, judicial, and administrative processes. Yet these human algorithms undeniably fail and are far from transparent. On an individual level, human decision-making suffers from memory limitations, fatigue, …

Using Ai To Reduce Performance Risk In U.S. Procurement, Jessica Tillipman

GW Law Faculty Publications & Other Works

In recent years, several U.S. government agencies have pioneered the use of artificial intelligence (AI) and other emerging technologies to improve the efficiency and accuracy of their "responsibility determinations" (reviews of, among other things, contractor representations and certifications, past performance history, civil and criminal settlements, exclusions (such as suspensions or debarments), and contract terminations). As federal agencies continue to think strategically about how to improve processes and reduce risk in their procurements, technology-driven solutions will play a critical role in this undertaking.

But What Is Personalized Law?, Sandra G. Mayson

All Faculty Scholarship

In Personalized Law: Different Rules for Different People, Omri Ben-Shahar and Ariel Porat undertake to ground a burgeoning field of legal thought. The book imagines and thoughtfully assesses an array of personalized legal rules, including individualized speed limits, fines calibrated to income, and medical disclosure requirements responsive to individual health profiles. Throughout, though, the conceptual parameters of “personalized law” remain elusive. It is clear that personalized law involves more data, more machine-learning, and more direct communication to individuals. But it is not clear how deep these changes go. Are they incremental—just today’s law with better tech—or do they represent …

From Negative To Positive Algorithm Rights, Cary Coglianese, Kat Hefter

All Faculty Scholarship

Artificial intelligence, or “AI,” is raising alarm bells. Advocates and scholars propose policies to constrain or even prohibit certain AI uses by governmental entities. These efforts to establish a negative right to be free from AI stem from an understandable motivation to protect the public from arbitrary, biased, or unjust applications of algorithms. This movement to enshrine protective rights follows a familiar pattern of suspicion that has accompanied the introduction of other technologies into governmental processes. Sometimes this initial suspicion of a new technology later transforms into widespread acceptance and even a demand for its use. In this paper, we …

Antitrust By Algorithm, Cary Coglianese, Alicia Lai

All Faculty Scholarship

Technological innovation is changing private markets around the world. New advances in digital technology have created new opportunities for subtle and evasive forms of anticompetitive behavior by private firms. But some of these same technological advances could also help antitrust regulators improve their performance in detecting and responding to unlawful private conduct. We foresee that the growing digital complexity of the marketplace will necessitate that antitrust authorities increasingly rely on machine-learning algorithms to oversee market behavior. In making this transition, authorities will need to meet several key institutional challenges—building organizational capacity, avoiding legal pitfalls, and establishing public trust—to ensure successful …

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 …