Few practices have higher stakes for human error than that of criminal justice. An oversight or delay can be the difference between life or death, justice or injustice. Artificial intelligence (AI) holds both opportunities and trials in the pursuit of safer communities and effective systems.

In December 2024, the United States Department of Justice (DOJ) released a report on Artificial Intelligence and Criminal Justice and identified areas of criminal justice where AI is applicable. Among them, identification and surveillance, forensic analysis, predictive policing, and risk assessment. AI’s ability to organize large amounts of data is well-suited to the detail-intensive but often resource-short nature of policing and many of the benefits enumerated in the report involve identifying and sorting images and other data. The report acknowledged that the technology could drastically improve efficiency and accuracy, but also posed challenges in areas such as civil liberties and discrimination.

Before the technology even hits the streets in the form of identification and risk assessment, police departments can use it to train officers and increase safer outcomes. At the University of Kentucky, researchers are developing methods of training that rely on AI to simulate de-escalation scenarios. Officers can learn to accurately navigate some of the most challenging and intense moments of their career in the safety of a virtual reality environment powered by AI.

In addition to crime prevention and solving, AI can support legal proceedings within the criminal justice system. Courts can use AI to streamline the administrative process that often contributes to the legal process taking far longer than it should. Louisiana 5th Circuit Judge Scott Schlegel recently highlighted the potential for reduced “failure to appear” rates and a more seamless legal experience through AI-powered agents: “When a party or their attorney receives a reminder about the upcoming hearing, the message could include a link to an AI assistant with access to the clerk of court’s system. With one click, they’d be able to ask case specific questions and get the information they need without having to dig around to find the number to the clerk’s office.” Thanks to the efficiency of AI, attorneys and litigants alike would have more time and resources to devote to the case at hand.

While the promise of AI for policing and the courts is evident, stakeholders must engage with difficult questions about fairness, accuracy, and accountability to ensure that the technology is used for good. The Council on Criminal Justice (CCJ) recently announced a National Task Force on Artificial Intelligence that aims to “meet the urgent need for credible guidance to help policymakers and practitioners navigate a complex and rapidly evolving landscape in ways that maximize benefits, minimize harms, and improve justice.” The task force will raise important questions in order to offer education, clarity, and resources to those within the criminal justice system.

Taking a learning approach (like CCJ’s), rather than a regulatory one, is the only feasible path forward. Efforts to keep AI away from industries and practices inevitably fall short-stunting progress and increasing confusion about standards and best practices. Similarly, avoiding honest discussions about the risks and training necessary to ensure good outcomes is shortsighted. Both the scales of justice and the system’s approach to AI must remain balanced.

Links to Learn More about AI

Work Smarter – Pelican Policy

Two Cheers for the AI Moratorium! | American Enterprise Institute – AEI

The Transformative Power of AI in Healthcare | U.S. Chamber of Commerce