On the way to robot judges
Developments in engineering and technology have led to the natural emergence of robots in social life, which began to independently perform the functions of regulators or participants in legal and social relations. For example, robots are already successfully performing the functions of transport management, power supply, and human life support. Robots save people in emergencies. Military robots (drones) have long been able to destroy people independently and en masse, not only during armed conflicts. Robots catch criminals, write news, and trade on stock exchanges. Artificial intelligence automatically diagnoses the sick. Can't robots realize the constitutional ideal of judges obeying only the law? They probably can because there are no higher bosses above them to whom they must abide. The artificial intelligence system eliminates the possibility of legal error. The robot knows only the law and acts according to it. A human lawyer is much more expensive than a robot lawyer. With a computer mind, legal cases can be handled very differently because only what the robot will accept as the law will matter. Of course, complete human trust in robots will not arise soon, but now it is becoming more and more difficult for people to compete with robots as equals, and maybe it is already beyond their power. And this is a legal reality that both society and its legislators must reckon with. In addition, computer intelligence has already been created that exceeds human capabilities. Moreover, there are cyber systems capable of self-learning and adapting to society's material and spiritual life. Artificial intelligence no longer only writes poetry, music, and paintings. It is already beating humans at chess and casinos and in many other ways. There are spore-readers programs that can easily outwrestle humans in legal debates. Programs that expose falsity and perjury are sensitive to people's emotional and psychological states. Programs have long been able to write and verify contracts made by humans. Artificial intelligence generates business and securities. It develops cryptocurrencies convertible into actual capital. It independently manages property rights. An algorithm has been created capable of accurately predicting state court decisions. Artificial modeling of judicial acts allowed robots to repeat the findings of the U.S. Supreme Court in 70% of cases and the European Court of Human Rights in 79% of cases. It means that it is only a short time before there are robots: judges, prosecutors, lawyers, etc. And they will be much faster and better able to solve the problems of lawmaking, judicial proceedings, and administration. They will be able to solve almost any legal case with mathematical precision. Unlike humans, they can impartially maintain the set parameters of objectivity and humanism. At the same time, they will not take bribes. A robot judge cannot be bribed, and human weaknesses do not matter to him. A robot judge is programmed to perform its functions in a quality manner by any legal means.
Artificial legal intelligence
There is an opinion that this is just a fantasy, which has nothing to do with real life because robots cannot take into account all the details, including the human factor. But most experts believe that society needs impartial, non-corrupt judges with as few mistakes as possible. They think it would be much better if an intelligent machine with no human side factors would create flawless laws and correct erroneous ones. Of course, if society becomes convinced that the same legal function can be performed by a robot as a human, that it is better than a human at the legal task, there will be no alternative to robots. Moreover, such flawless systems have already been created, and they work. In Munich, a computer program called Elterngeld, Parents' Money, has been designed to make decisions on child support claims. It has been making financial and legal choices transparent for several years now. In Finland, the TrademarkNow legal robot handles trademark claims. It operates a linked database of U.S. and European registrars and makes error-free decisions. The U.S. law firm Baker & Hostetler has used the ROSS robot with artificial intelligence to pick up precedents in specific bankruptcy, tax, and health insurance cases since 2016. What a human lawyer spends hours on, a robot lawyer takes moments. The DoNotPay robot developed at Stanford University disputes fines to car owners. In two years, the robot auto lawyer has served over 250,000 customers, most of whom have avoided penalties. JPMorgan's COIN program has automated its lawyers' work analyzing credit contracts. Previously, this work required hundreds of thousands of person-hours of labor annually. Now cases are solved in seconds. Thus, the main goal is to move from the jurisprudence of human error to impeccable and accurate legality.
"Smart" contracts
Be that as it may, circumstances are conducive to using AI in jurisprudence. In particular, mathematical methods and algorithms have formed so-called smart contracts. These computer protocols map or verify contracts or provide technical support for negotiating or enforcing a contract. "Smart contracts increase the security of legal relationships over traditional contract law, reducing transaction costs. A computer program enforces a contract built into the code with smart contracts.
"Predictive justice" programs
In addition, LegalTech programs that predict the outcome of litigation, including potential compensation, are now emerging. LegalTech offers many benefits: automating repetitive tasks for lawyers, distractions, reducing litigation risk, etc. The purpose of developing and implementing such tools is to automate repetitive tasks so that lawyers can focus on more priority functions, first of all, consulting clients. Thus, a new concept appeared in European countries - "predictive justice" (PJ). PP relies on AI, giving lawyers the tools to focus on more complex tasks. While traditional justice tries to fulfill its mission of deciding a case within a reasonable time, PP goes further. It provides algorithms for analyzing a vast number of situations in a short period, which allow you to anticipate the outcome of the dispute or at least assess the chances of success. PP allows: to choose a correct way of defense, choose the most appropriate arguments, estimate the expected amount of compensation, etc. Thus, it is not a question of justice itself but only of analytical tools that would make it possible to predict future decisions in disputes similar to those analyzed. The term "PP" itself comes from the Anglo-Saxons. In the U.S., state-level software has long been used, particularly in criminal law, as well as databases for enforcement of regulations, document classification, etc. For example, the Los Angeles Police Department has adopted a predictive policing system called PredPol. It uses data on the behavior of individuals and more effectively directs police efforts to fight crime. U.S. judges often use programs to assess a suspect's likelihood of reoffending, such as Correctional Offender Management Profiling for Alternative Sanctions (COMPAS). True, a recent study conducted by University College London showed that computers could predict the trial's outcome with an accuracy of about 80%. But how acceptable is that figure? Once people start using artificial intelligence, and they expect the outcome to be predicted by machines, what will judges do then? And if judges don't perform according to the expectations of artificial intelligence, it will become a problem for the court. And as an outcome, there will be a question about the necessity of its existence as an institution.
International norms on "predictive justice"
International norms on AI in justice have already been developed and enacted. Thus, on December 4, 2018. European Commission for the Efficiency of Justice (CEPEJ) of the Council of Europe adopted the first document - "Ethical principles concerning the use of artificial intelligence in judicial systems." This document is a guide for lawmakers and justice professionals. The use of AI in justice must be carried out responsibly that is consistent with the fundamental rights guaranteed in the European Convention for the Protection of Human Rights and the Council of Europe Convention on the Protection of Personal Data. The CEPEJ has identified the following principles to be respected in the justice field:
- Respect for fundamental rights: ensuring compliance and implementation of AI tools and services with fundamental rights,
- non-discrimination,
- quality and security regarding the processing of judgments and data in a secure technological environment,
- transparency, impartiality, and fairness,
- ensuring that data processing methods are accessible and understandable,
- "under user control," etc. On April 10, 2018. 25 European countries have endorsed a Declaration of Cooperation in AI. Several states have already announced national initiatives in this area and declared their intention to adopt a pan-European approach to solving the problem. Of course, all of these provisions are not yet EU legal norms. Legal norms on the subject will be adopted soon. Their purpose is not to create a robot judge capable of making decisions without humans. Their goal is to strengthen the analytical capacity of the best lawyers. The states of Europe are aware of the new possibilities of PP and are actively developing relevant programs at the national level.
National "predictive justice" programs
For example, since 2015, the Netherlands has had Recht-wijzer, the Conciliation and Mediation Platform, which covers all proceedings involving individuals, including rent, and family disputes. In March 2018. The Justice and Security Commission of the Lower House of the Dutch Parliament held a roundtable on AI in Justice, which discussed the need to regulate PP.
France is a recognized leader in the implementation of AI in Europe. As we know, France has more than 50 codes, about 10,500 laws, 127,000 decrees and other regulations, and millions of court decisions.
On October 7, 2016, the law on the Digital Republic was adopted, which detailed the procedures for exchanging and accessing public information.
To achieve successful PP projects, two conditions are necessary. First, quality data in large quantities. The 2016 Digital Republic Act removed this obstacle. Second, technological developments have made powerful algorithms possible by recent technological developments. With the development of new technologies, PP makes it easier to process two million court decisions a year. In France, several PP programs are already in place. For example,
- justice.fr is designed for lawyers and allows them to identify the competent court and download the appropriate forms for filing an application/complaint.
- Sagace allows the parties to get acquainted with the summary information of the court case.
- Consultation Avocats is a national platform for counsel (by appointment, phone, or email), whose services are then covered by a fee agreement.
- JuriCA and JuriNET, the case law database of the civil and commercial appeal courts (JuriCA), and the court of cassation for all cases (JuriNET).
There are two types of players in the PP market in France: classic legal publishers and LegalTech startups. The evolution of LegalTech is facilitated by a combination of favorable factors: the digital revolution, changes in mentality, and the emergence of new technologies.
In France, they believe that the use of PP is especially promising in social law. In particular, entrepreneurs without a legal background regularly face various difficulties in the area of labor law, especially when drafting contracts and resolving disputes with employees. With the introduction of President Macron's decree prohibiting dismissal without serious justification, the use of PP becomes particularly important. The Droit social database contains 470,000 cases concerning severance payments, compensation for work accidents, etc. Various search criteria are available (such as the type of contract, the type of injury, the type of accident, etc.). Different search criteria are general (company characteristics, employees, classes, amounts of allowances and damages, etc.), making life much easier for employees and employers alike.
The Future of "Predictable Justice"
In recent years, powerful forces have changed the market for legal advice provided to corporations. The new reality is characterized by falling prices, transforming previously homogeneous services into different types, and the growth of outsourcing legal procedures. In France, for example, not all companies have a legal service. Most companies are small and medium-sized and do not have a residual number of financial resources; it is not always available to use the services of a lawyer. Therefore, for them, PP programs are beneficial.
The benefits of PP for the courts
Information technology can provide tools to make judges' decisions easier or better guide the choices of citizens and legal representatives. For example, a judge who has to make a decision on a complex case can inquire about relevant case law. PP programs provide analysis of issues already decided and identify similar situations. Appellate and supervisory trends can also be easily detected. Using algorithms, plaintiffs can learn about the success of similar cases and adjust their strategies if necessary. Digitization of legal data represents another megatrend that is transforming LegalTech's workflows and business models. The amount of data used in legal advice has increased exponentially, as seen in many other industries. For example, in cases involving Enron, Lehman Brothers, and Volkswagen, lawyers analyzed and interpreted hundreds of thousands of emails and other documents to establish the facts. In the future, the ability to verify, research, and interpret unprecedented amounts of data will become as crucial to the success of law firms as the "art" of giving legal advice. Various legal technologies have emerged to digitize and automate these and other legal activities. In fact, the results of a survey of partners and legal technology providers conducted by The Boston Consulting Group suggest that LegalTech solutions can provide up to 30-50% of the tasks performed by paralegals today. Yet LegalTech has been adopted by fewer law firms than might be expected. Many lawyers still do data analysis using the traditional approach: manually going through documents one by one. On the one hand, the current revenue model based on billable hours and profit-sharing agreements governing partnerships provides little incentive to invest in new technology. On the other hand, many law firms are wary of adopting solutions still in development, requiring significant operational changes and investment in training. Regardless of the reasons for the not-so-high adoption rate of PP technologies, law firms that fail to take advantage of them or adapt to the new realities they present risk losing corporate clients to more forward-thinking practices and may go bankrupt. Suppliers of legitimate technology products will replace them. The development of IT has led to more human interaction everywhere. As a result, public services, including the justice service, have sought to take advantage of new digital tools to redefine how they communicate with their users. Providing additional information and improving its delivery are two goals that court program operators have found possible at lower deployment costs thanks to AI. The focus is on gaining two benefits: improving service delivery quality while controlling the justice system's operating costs. Thus, lawyers on their own (those who do not recognize technology), just as robots themselves (created to replace humans), will never have the same value as humans and machines together. It is the key to solving the whole problem. Government agencies, lawyers, and law firms can no longer afford to ignore PP technology. To ensure their competitiveness and survivability, all law enforcement entities must be prepared to take advantage of the opportunities provided by PP technology.