The Chapman Law Review is proud to publish Aubrey A. Butler’s note: Fingerprints of Injustice: The Truth Behind Artificial Intelligence and Algorithmic-Driven Evidence. Below, you will find the abstract from the article.

Fingerprints of Injustice: The Truth Behind Artificial Intelligence and Algorithmic-Driven Evidence

By Aubrey A. Butler

Abstract

Artificial intelligence (AI) has steadily grown in prominence, reaching into nearly every aspect of daily life, and the legal system is not immune from its influence. As various forms of machine evidence have been admitted in court, there has been an accompanying rise in concerns from researchers, judges, and defense attorneys as to the trustworthiness and reliability of this new category of evidence.

Though AI and machine evidence takes various forms, this Note focuses on some of the most prominent programs being used in courtrooms today, namely the probabilistic genotyping software TrueAllele and the recidivism rate prediction software COMPAS. The creators of these programs guarantee their accuracy, yet several studies have demonstrated proven defects in their operation—defects which cannot be fully tested and resolved due to the proprietary or “black box” nature of the underlying source code.

Several solutions have been suggested for how to approach machine evidence moving forward, but this Note posits a new mechanism which has not been previously addressed: the creation of a new federal agency focused on AI within the United States with a department wholly dedicated to computer-driven evidence in the legal system. This agency would be able to analyze machine evidence and send out scientific advisors to courts to counsel judges about the potential dangers of this form of evidence in a way that is not possible under the current system.