Recently, a ruling by the European Court of Justice asserted that Facebook could be made to delete content across the border or worldwide. The verdict was in line with the European Union’s internet laws. Now, Facebook laments that the decision sets a very poor precedent for future and other international organizations. Some feel that it is an overarching authority exerted by the court and many others adore the ruling. The point is that their fellow beings always consider a human panel of judges as incapable of objective decision making.
A strong case for AI in courtrooms
One of the primary reasons people are not satisfied with the court rulings at lower levels is that both parties hope that a different human judge may see things differently. The hope leads to more and more appeals to the apex courts in various countries. The final consequence of this upward push is that the cases go on pending. For example, over 30 million cases are pending for more than 25 decades in the Indian courts. As the old saying goes “Justice delayed is justice denied.” Hence, slower delivery is indeed a case of injustice.
A second reason is the human factor. In a study conducted in Israel in 2010, it was found that stress of work had a significant impact on the quality of judgement. The research shows that judges while hearing parole pleas of convicts were more forbearing in the morning than in the afternoon, and in the first session than in the second. It indicates the effect of stress on their perceptive power.
While justice delayed is a form of injustice, so is a judgement given in a tight spot. All this creates a strong case for non-human intelligence. And in this era of technological progress, Artificial Intelligence offers a different pathway.
The AI alternative
While considering AI as an alternative for faster judgement, two prominent approaches emerge. One method involves AI assisting the human judge, and the second one is through the separation of AI and human judge.
In the first approach, AI will be assisting the human judge in getting to case details faster. Here the AI will be playing an advisory role, and the final discretion on whether to accept the logical conclusions of the AI will rest with the human judge. The second method involves delegating a part of work to the AI which and another part to human judge.
The first methodology stresses the transparency aspect at the cost of performance and the second aspect focuses on performance. An issue which still needs consensus for the use of AI in courts is the balance of transparency and performance. The extent to which we can trust the reasoning of the machine, the possibility of error and the ruling on delicate value-based issues is something that academicians are still debating on.
The metaphors of AI
When we consider the involvement of AI can take in a courtroom, the metaphors or the particular name that we give to AI’s job function is very relevant. The importance arises from the necessity to define the degree of powers that we can delegate to the AI element. The metaphor will be pivotal in designing the human-technology interaction in courtrooms. It will also reflect the work that AI will perform in the judiciary.
Here we discuss some of the metaphors that AI can take up to perform specific roles. Each role that AI plays expedites the efficiency in the execution of the task and increases the overall pace of disposing cases.
AI as the court librarian
A court librarian as a human element is as old as a traditional form of the judiciary. In the earlier days, when technology was not so advanced, the judge used to assign to the librarian the task of finding particular literature from the whole list of judgements and presets.
AI can work wonders in this role because a human agent who has to peruse through various files and look for particular keywords and tags is going to take considerable effort and time. The intelligent use of algorithms and machine learning can make the AI- Librarian a smart asset for the judiciary.
However, the path is not all clear. There are significant hurdles to overcome the ethical questions of using machines in court like transparency, the delegation of judicial power etc. In the earlier days, the librarian used to fetch the judge the literature with an intelligence of his own. There was also a room for discussion between the judge and librarian to exchange the criterion that he followed in zeroing on printed materials.
Luminance is an AI-based tool used in the United Kingdom, and it uses machine learning to study the interactions between the lawyers and documents and rewire algorithms. Luminance is an example of what AI can do in text-analysis.
AI as advocate generals
The definition of Advocate General (AGs) may vary from one country to another. Still, considering the European Court of Justice mandate, an AG is someone who assists the court in responding to a particular case. An AG looks for similarities in a suit and independently offers his or her suggestion for a possible course of action if a situation doesn’t raise a new interpretation of the law.
An Intelligent Advocate General is ethically pleasing to the skeptics as well, because the role is more of an assistive nature. The judge can quickly look through the report created by AI and has a final control on the judgment as well as the logic that led to the judgement as sources are clearly made available.
AI as judicial officials with limited powers
AI as a separate judicial officer is an ultimate solution with the onus of approaching them lying on the applicants. The petitioner can decide whether to go for an AI or human aid. The provision of appeal to a human judge can be provided to reinspect the logic of the AI official. Estonia is already on the way to create a robot-judge to clear its long list of illegal backlogs. A clear definition of conflicts and resolution can help in making millions of pending cases that require mere consideration of plain facts.
The AI is now permeating all boundaries to find a way into our daily life. Judiciary is no exception as the predictive crime analysis tools like Preditice in France and HART (Harm Assessment Risk Tool). The mistrust in some section of society towards AI will evaporate as they will see the experience the efficacy of them in other aspects of life. The future is bright for AI in judiciary and time will indeed decide its role in the system.