AI is already making dozens of decisions for us every day about what to read, watch or buy. In the future, it could replace judges. A team of British and American computer scientists have developed a system that accurately predicted rulings by human judges in nearly 80% of the real-life cases it examined. The machine was trained with hundreds of court rulings and key words. The algorithm cannot yet make nuanced decisions in complex settings better than humans, although futurologists predict that too will change in a few decades. “‘Explainable AI,’ which is machine learning that can explain the steps it took to achieve a decision, is getting underway,” Amy Zalman, professor of strategic foresight at Georgetown University, tells OZY.

There’s more. From the questionable role of private forensic examiners to voice-recognition software designed to help police departments prioritize calls, new technologies are finding their way into every aspect of the judicial process. Even data gathered by devices tracking our steps and heartbeats has been used as evidence in court. In China, authorities are taking things even further. They are believed to be building a massive DNA database, which could easily, and dangerously, be used for policing. And it’s not just justice that is getting an upgrade. Crime is also evolving thanks to new, increasingly popular technologies (think cybercrime and drones that fly drugs across the U.S. border, for example).

One of the most radical changes to the justice system has meant making it much more accessible. The ability to join a court procedure from almost anywhere is particularly welcome news for those who cannot afford to take time off work. Some lawyers, however, miss the face-to-face action. They say effective cross-examination is impossible to replicate without everybody being in the same room. “The best part of an oral argument is what I call jazz: ​​there is improvisation, there is interruption, and the sense that you might change someone’s mind in the heat of the moment,” Kathleen Sullivan, a partner at Quinn Emanuel Urquhart & Sullivan, told Harvard Law Today. “Instead of jazz, you have something highly formalized, a kind of orchestrated chamber music.” Another potential downside to the new “Zoom courts”? If you’re not careful, you might turn into a cat.

A major problem with deploying algorithms in the service of the public good is that they can be racist and sexist, often favoring Caucasian-looking faces. Why? One reason is that they replicate the same biases apparent in the data from which they are fed. Another, Amy Zalman explains, is that these technologies are built by developers who, for the most part, are white and male. “Our social systems will always be created first by people,” she tells OZY, adding that ultimately, humans tell the machines what to think. But she remains optimistic that “there are ways in which technology can be helpful. There are experiments to blind hirers to what potential employees look like, for example.”

Your phone is already your office, cinema and personal spy. You know what else is in there? A lawyer. The DoNotPay app is a “personal attorney” that helps people with issues ranging from a contested parking ticket to an unfair banking fee. Created in 2018 by U.K.-based tech entrepreneur Joshua Browder, the app generates letters users can send to authorities, for a small monthly fee. That’s not all. In U.S. and U.K. law firms, so-called “robot lawyers” are increasingly being used to skim through large volumes of files searching for relevant information. These modern-day paralegals could change the face of the profession, compelling lawyers to swap law books for AI manuals.

There’s a prison revolution taking place, although not the type you might imagine. This one involves the use of voice and video analytics, applying real-time smart surveillance systems to predict antisocial conduct from both inmates and correctional officers. In the U.S., for example, Congress is pushing for more research into how AI could better analyze inmates’ phone calls. The idea is for AI to help identify words and phrases and even evaluate someone’s tone of voice, all of which can assist crime investigators. Civil rights advocates, however, say these technologies are prone to disproportionate error rates when interpreting the voices of people of color.

New tech could also soon replace brick and mortar jails. New generation GPS locators would allow authorities to keep a closer eye on offenders while allowing them to live in environments that better support rehabilitation (think low crime areas). Imagine a prison that is less violent and cheaper to run. “More forward-looking jurisdictions will increasingly integrate online learning, skills training and mental health resources, transforming prisons into true reform institutions that can transform criminals into valued members of society,” David Tal, president of Toronto-based Quantumrun Foresight, tells OZY.

In addition to the infrastructural changes set to take hold, the concept and understanding of punishment itself will radically change. Futurologists say scientists working with attorneys and lawyers will likely be able to separate the mental from the physical punishment. This could involve using brain manipulation techniques to erase the traumatic, violent and criminal memories that lead prisoners to engage in criminal activity, Tal explains. Looking further into the future, he says governments might one day sync prisoners’ brains to a virtual prison while keeping their bodies in a stasis tube, Matrix-style.

A line of thinking among futurologists suggests technology can, in fact, democratize access to justice to a degree like never before. Scientific advances such as AI will help people better understand their rights and navigate complex legal systems that at present are only accessible to lawyers. Also, robot paralegals can help make litigation cheaper and more agile, resolving the massive backlog of cases many countries struggle under (along with high rates of pretrial detention). Tal goes even further: “Police departments and court systems can use various AI solutions to automate bureaucratic functions to reduce costs and free up officer time for street patrols and community engagement.”

But it’s too soon to celebrate the legal revolution. In many countries, the kind of technology required to implement these changes will be prohibitively expensive for decades to come. Also, the inherent racial and gender biases that attach to today’s algorithms demand a more diverse tech industry — and that too will take time, Zalman says. But even in countries where resources aren’t an issue, a major struggle going forward will be the need for law schools to dramatically change how they train the lawyers of tomorrow so they are better equipped to deploy these new technologies.

In Europe, authorities have adopted a charter to incorporate technologies such as AI into their criminal justice systems. The Council of Europe, which is in charge of the European Court of Human Rights, warned against the use of predictive justice. The body argues that, aside from racial bias algorithms, new technologies are also unable to account for the context in which events take place. Before embracing data-driven legal solutions, the organization believes its important to ensure users fully understand these new systems. However, despite all the fears associated with the technology use, the charter was accepted.