Artificial intelligence (“AI”), and machine learning in particular, promises lawmakers greater specificity and fewer errors. Algorithmic lawmaking and judging will leverage models built from large stores of data that permit the creation and application of finely tuned rules. AI is therefore regarded as something that will bring about a movement from standards towards rules. Drawing on contemporary data science, this Article shows that machine learning is less impressive when the past is unlike the future, as it is whenever new variables appear over time. In the absence of regularities, machine learning loses its advantage and, as a result, looser standards can become superior to rules. We apply this insight to bail and sentencing decisions, as well as familiar corporate and contract law rules. More generally, we show that a Human-AI combination can be superior to AI acting alone. Just as today’s judges overrule errors and outmoded precedent, tommorrow’s lawmakers will sensibly overrule AI in legal domains where the challenges of measurement are present. When measurement is straightforward and prediction is accurate, rules will prevail. When empirical limitations such as overfit, Simpson’s Paradox, and omitted variables make measurement difficult, AI should be trusted less and law should give way to standards. We introduce readers to the phenomenon of reversal paradoxes, and we suggest that in law, where huge data sets are rare, AI should not be expected to outperform humans. But more generally, where empirical limitations are likely, including overfit and omitted variables, rules should be trusted less, and law should give way to standards.

In the wake of widespread revelations about sexual abuse by Harvey Weinstein, Larry Nassar, and others, the United States is reckoning with the past and present and searching for the means to prevent and punish such offenses in the future. The scourge of sexual crimes goes far beyond instances perpetrated by powerful men; this misconduct is rampant throughout the country. In some of these cases, third parties knew about the abuse and did not try to intervene. Scrutiny of—and the response to—such bystanderism is increasing, including in the legal world.
In order to align law and society more closely with morality, this Article proposes a more holistic, aggressive approach to prompt involvement by third parties who are aware of specific instances of sexual crimes in the United States. This Article begins by documenting the contemporary scope of sexual crimes in the United States and the crucial role bystanders play in facilitating them.
The Article next provides an overview and assessment of “Bad Samaritan laws”: statutes that impose a legal duty to assist others in peril through intervening directly (also known as the “duty to rescue”) or notifying authorities (also known as the “duty to report”). Such laws exist in dozens of foreign countries and, to varying degrees, in twenty-nine U.S. states, Puerto Rico, U.S. federal law, and international law. The author has assembled the most comprehensive global database of Bad Samaritan laws, which provides an important corrective to other scholars’ mistaken claims about the rarity of such statutes, particularly in the United States. Despite how widespread these laws are in the United States, violations are seldom, if ever, charged or successfully prosecuted.
Drawing on historical research, trial transcripts, and interviews with prosecutors, judges, investigators, and “upstanders” (people who intervene to help others in need), the Article then describes four prominent cases in the United States involving witnesses to sexual crimes. Each case provides insight into the range of conduct of both bystanders and upstanders.
Because not all such actors are equal, grouping them together under the general categories of “bystanders” and “upstanders” obscures distinct roles, duties, and culpability for violating those duties. Drawing on the case studies, this Article thus presents original typologies of bystanders (including eleven categories or sub-categories), upstanders (including seven categories), and both kinds of actors (including four categories), which introduce greater nuance into these classifications and this Article’s proposed range of legal (and moral) responsibilities. These typologies are designed to maximize generalizability to crimes and crises beyond sexual abuse.
Finally, the Article prescribes a new approach to the duty to report on sexual abuse and possibly other crimes and crises through implementing a combination of negative incentives (“sticks”) and positive incentives (“carrots”) for third parties. These recommendations benefit from interviews with sexual violence prevention professionals, police, legislators, and social media policy counsel. Legal prescriptions draw on this Article’s typologies and concern strengthening, spreading, and standardizing duty-to-report laws at the state and territory levels; introducing the first general legal duty to report sexual crimes and possibly other offenses (such as human trafficking) at the federal level; exempting from liability one of the two main bystander categories the Article proposes (“excused bystanders”) and each of its six sub-categories (survivors, “confidants,” “unaware bystanders,” children, “endangered bystanders,” and “self-incriminators”); actually charging the other main bystander category the Article proposes (“unexcused bystanders”) and each of its three sub-categories (“abstainers,” “engagers,” and “enablers”) with violations of duty-to-report laws or leveraging these statutes to obtain testimony from such actors; and more consistently charging “enablers” with alternative or additional crimes, such as accomplice liability. Social prescriptions draw on models and lessons from domestic and foreign contexts and also this Article’s typologies to recommend, among other initiatives, raising public awareness of duty-to-report laws and creating what the Article calls “upstander commissions” to identify and “upstander prizes” to honor a category of upstanders the Article proposes (“corroborated upstanders”), including for their efforts to mitigate sexual crimes. A combination of these carrots and sticks could prompt would-be bystanders to act instead as upstanders and help stem the sexual crime epidemic.

For half a century at least, the several states of the United States have taken a liberal attitude toward the recognition and enforcement of foreign country money judgments. The U.S. Supreme Court invoked the “grace” of sovereign nations to justify a restrictive approach to the recognition of judgments in the famous case of Hilton v. Guyot. The New York Court of Appeals laid out a more generous approach based in the vindication of private rights. Simply put, private rights won. In 1962, the Uniform Law Commission promulgated the Uniform Foreign Money-Judgments Recognition Act, which codified a liberal approach to the cross-border circulation of money judgments. The many U.S. states that adopted the uniform act were trying to lead by example. The hope was that, if they accepted incoming judgments, judgments exported to the rest of the world would be accepted, recognized, and enforced. For decades, this effort was regarded as a failure. The European Union continued to draw a sharp distinction between E.U. judgments and U.S. judgments—though acceptance of U.S. judgments by E.U. member states crept up over time. Some of the world’s largest economies—most notably, China—outright rejected recognition of U.S. money judgments.

Change has been recent and dramatic. In 2017, a Chinese court recognized and enforced a U.S. money judgement for the first time. Chinese law requires reciprocity between nations in order to recognize a foreign money judgment. The United States has no reciprocal judgment recognition treaty with any country. A U.S. district court recognized and enforced a Chinese judgment in 2009. This “reciprocity in fact” was sufficient for a Chinese court. A few months later, China announced that it would sign The Hague Convention on Choice of Court Agreements (“COCA”), obligating Chinese courts to recognize and enforce judgments rendered under a choice of court clause selecting the courts of any contracting state. The COCA has already entered into force between the European Union, Mexico, and Singapore. The United States has signed, but not ratified, the agreement. Meanwhile, The Hague Judgments Project gathers steam to require the free circulation of judgments arising in all but a few contexts. The drivers of this apparent convergence are obscure and likely diverse. This Article will analyze the causes of this recent, dramatic shift and will attempt to assess the likelihood of further convergence.

This Article explores the divergence in law and convergence in economics in dealing with harms and benefits. While tort law usually makes the injurer internalize wrongful harms through damages, restitution law does not enable the benefactor to internalize the benefits she confers on others without their request. In both harm and benefit cases, however, internalization seems to make economic sense for the same reason: injurers and benefactors alike will behave efficiently if they internalize the externalities that they create. The Article’s main goal is to develop eight liability rules for harm and benefit cases and to point out the symmetry between the rules relating to harms and the rules relating to benefits. It also provides an explanation for the legal divergence between tort law and restitution law and makes the claim that the gap between these two fields should be narrowed. Finally, the Article relates these eight rules to the main relevant categories of harm and benefit cases in positive law and appraises their advantages and disadvantages.

In this Article, we study rules that solve the conflict between the original owner and an innocent buyer of a stolen or embezzled good. These rules balance the protection of the original owner’s property and the buyer’s reliance on contractual exchange, thereby addressing a fundamental legal and economic trade-off. Our analysis is based on a unique, hand-collected dataset on the rules in force in 126 countries. Using this data, we document and explain two conflicting trends. There is a large amount of first-order divergence: both rules that apply to stolen goods and those that apply to embezzled goods vary widely across countries. Yet, there is also remarkable second-order convergence: virtually all legal systems protect the innocent buyer more strongly if the good was embezzled (rather than stolen) and if she purchased it in an open market, at an auction, or from a professional seller (as opposed to a private sale). We show that, while divergence is attributable to varying cultural values, convergence can be rationalized using a classic functional approach: these rules harmonize the owner’s incentives to protect property and the buyer’s incentives to inquire about title.