Publications

For too many people experiencing acute mental illness, cries for help bring police and handcuffs rather than compassionate medical intervention. But this doesn’t have to be the case. This report examines the laws across the U.S. related to law enforcement custody and transportation under involuntary civil commitment, when alternative transport is permitted, and opportunities to reduce the role of law enforcement in these situations when possible.

Since 1989, 75 people have been exonerated in North Carolina. Collectively, these exonerees lost 963 years to incarceration for crimes they did not commit. Despite a wave of reforms in the state in the early 2000s designed to prevent wrongful convictions, 11 people have been convicted and later exonerated since those reforms. This report documents the reforms instituted in the early 2000s, examines the 11 known wrongful convictions since that time, and makes recommendations for further reform to prevent wrongful convictions and improve the paths to exoneration for those who have been wrongfully convicted.

Most jurisdictions in the United States now record interrogations, including all federal law enforcement agencies, thirty states, and the District of Columbia. This memo provides updated information as of August 2024 about the state-level adoption of electronic recording requirements, through statutes, court rulings and rules, and police policies.

This amicus brief was filed in collaboration with the Innocence Project and Greenberg Traurig, LLP. Mr. Bolin has served 28 years on death row, and his conviction was based on several questionable evidentiary methods. We argue that the eyewitness testimony in his case was highly suggestive and unreliable.

Brandon L. Garrett, Alicia Carriquiry, Karen Kafadar, Robin Mejia, Cynthia Rudin, Nicholas Scurich, and Hal Stern wrote this memo in response to NIJ’s request for public input regarding the response to Executive Order 14110, regarding the use of artificial intelligence (AI) in the criminal justice system. They wrote that in high-risk settings like the criminal justice system, AI models and underlying data must be adequately, transparently, and independently tested and that it must be interpretable. They wrote that far more can and should be done to apply and robustly protect the existing Bill of Rights in the U.S. Constitution as it should apply to uses by government of AI in the criminal system, particularly when AI is used to provide evidence in investigations and trials.

The number of incarcerated seniors has tripled in the last 20 years, and that number is only expected to grow. By 2030, U.S. prisons will incarcerate 400,000 seniors – 1/3 of the total prison population. Estimates vary but more than half of incarcerated seniors may develop dementia by that point. This policy brief by Megan Moore and Angie Weis Gammell makes recommendations on best practices for care and screening of incarcerated people with dementia as well as for their medical release. 

We partnered with Kelly Woodruff at the Complex Appellate Litigation Group to file a an amicus brief in the California Court of Appeals, challenging the reliability of firearm/toolmark evidence. 

For over a hundred years, firearms examiners have testified that they can conclusively identify the source of a bullet or cartridge case. In recent years, however, research scientists have called into question the validity and reliability of such testimony. In this article published in the Southern California Law Review, the authors detail over a century of case law and examine how judges have engaged with the changing practice and scientific understanding of firearms comparison evidence. The more-than-a-century-long arc of judicial review of firearms evidence in the United States suggests that, over time, scientific research can displace tradition and precedent to improve the quality of justice. By Brandon Garrett, Eric Tucker and Nicholas Scurich.

As governments and corporations use AI more pervasively, one of the most troubling trends is that developers so often design it to be a “black box.” Designers create AI models too complex for people to understand or they conceal how AI functions. Both champions and critics of AI, however, mistakenly assume that we inevitably face a trade-off: black box AI may be incomprehensible, but it performs more accurately. But that is not so. In this article published in the Cornell Law Review, authors Brandon Garrett and Cynthia Rudin question the basis for this assumption, which has so powerfully affected judges, policymakers, and academics. They describe a mature body of computer science research showing how “glass box” AI—designed to be fully interpretable by people—can be more accurate than the black box alternatives. Indeed, black box AI performs predictably worse in settings like the criminal system. Thus, Garrett and Rudin argue for national and local regulation to safeguard, in all criminal cases, the right to glass box AI. By Brandon Garrett and Cynthia Rudin.

On November 21, 2019, U.S. District Court Chief Judge Lee Rosenthal approved the Consent Decree in ODonnell v. Harris County. The new rules laid out in this consent decree require prompt release on unsecured bonds for the vast majority of people arrested for misdemeanor offenses. The Monitor and Deputy Monitor, Brandon Garrett and Sandra Guerra Thompson, are responsible for monitoring the progress in Harris County. This fact sheet shares key findings over the first four years after reform, including changes in release rates, racial disparities, and effects on public safety.