This article examines the phenomenon of false confessions in cases in which individuals were falsely accused of murdering close family members. Drawing on a dataset of twenty-two proven false confession cases, we document how grief, trauma, and coercive interrogation practices intersect to produce highly counterintuitive admissions and narratives of guilt. The study situates these cases within the broader literature on police interrogation and false confessions, highlighting situational risk factors such as lengthy custodial interrogation, sleep deprivation, false evidence ploys, and threats and promises, alongside dispositional vulnerabilities such as youth, mental illness, and trauma exposure. Unlike prior aggregated studies of proven false confessions, this analysis focuses specifically on the unique psychological burden faced by suspects who are grieving the violent death of a loved one and who are often interrogated within hours of the discovery of the crime. Many were misclassified by police as guilty due to their grief reactions, psychologically coerced into falsely confessing and fed crime scene facts and details that were repeated back in their false confessions. Our qualitative and Quantitative data analysis reveal recurring themes that have been well-documented and found in coerced-compliant and coerced-persuaded false confessions. The authors conclude with policy recommendations—including mandatory electronic recording of interrogations—and a call for trauma-informed investigative practices to prevent wrongful convictions rooted in coerced confessions from grief-ridden individuals in a uniquely vulnerable situation.

Automated, data-driven decision-making can create unfair outcomes and lead to discrimination. This Article considers a relatively ubiquitous part of modern life that is increasingly automated: the criminal background check. This Article, with contributions at the intersection of law and technology, antidiscrimination and consumer protection law, and sociological theory, makes the central claim that criminal record data is characterized by function creep—the unintended use of data for another purpose—which leads to a specific set of harms. This Article makes three central contributions. First, it offers an empirical assessment of contemporary, data-driven background screening using data based on in-depth interviews and systematic analyses of 104 New Jersey residents’ criminal records from both public and private sectors. Specifically, people in the study face three crucial data issues: 1) incorrect data, 2) misleading data, and 3) unknowable data. Second, the Article establishes the mechanisms of discriminatory harms as rooted in function creep, bridging scholarship in law, policy, and social science. Finally, the Article outlines how existing regulatory approaches fail those who are harmed and exacerbate the discriminatory and punishment-related harms of the criminal legal system. Overall, the Article establishes the fundamental problems that emerge when information created for processing cases through the criminal legal system is used to create background reports and predictive risk scores for profit. The rise of algorithmic data matching and automated decision-making further conceals the source data, making it increasingly difficult for people to gain access to, understand, or challenge their background check. This led many respondents in the study to withdraw from challenging these problems altogether. At the same time, both the agencies that have created criminal record data and the companies that commercialize it evade accountability. The Article concludes by suggesting specific areas of federal and state-level reform but cautions that such a focus overlooks the fundamental problem of using poor quality and often misleading criminal legal system data to assess people’s suitability for a job, an apartment, or full participation in society.