Robots To Deliver Judgement Very Soon

Dramatic changes are brewing in the global system of justice administration: the algorithms are being developed to expedite trials and anticipate (in order to prevent) crimes and criminal intentions.

If the experiment in the field of AI is successful, it can be assumed that robots will take the place of the traditional judge in the foreseeable future, and the potential criminal will be seized even before the crime is committed to be sentenced by the robot judge for criminal intent. Such a perspective will raise a number of important moral and ethical issues: the possible restriction (or reduction) of the fundamental rights and freedoms of man and citizen and the fairness of the trial and sentence.

What seemed like an unattainable virtual sphere yesterday, became a reality today. In particular, in Estonia, which is a leader in the field of e-government and e-residence, the judicial practice already applies the algorithm in administrative claims with damages of up to 7,000 euros.

This is called the “automated decision system”; it is able to examine the documentation submitted by the parties of the trial in electronic format, compare it with the current regulatory legal acts and make a decision on the case. In the event of appeal against a decision made by a system based on an algorithm, the protesting party may resort to an appeal, the decision on which will already be made by a human judge. The purpose of this approach is to accelerate the process of adjudication and reduce the burden on the clerical staff and representatives of the judiciary.

A positive aspect of robots and algorithms is comparative law enforcement practice: not a single human judge is able to conduct a comparative analysis of all precedents in judicial practice for a similar lawsuit. This kind of process and sentence are ideal for those who prefer a cold, balanced judgment, without emotion, anger, and addiction. Opponents, however, believe that in the event of an algorithm failure, a robotic court decision can lead to irreparable consequences and that a human judge should always assist the robot. This question has one more moral aspect: is the biblical saying “do not judge, let not you be judged” apply to a robot judge? Indeed, while penalized judges are judged, will a guilty robot judge be judged?

For several years now, in the United States, courts have used algorithms developed by the private company Compas to determine the amount of bail and, even more ambiguous from the point of view of public perception, to determine the degree of risk of recidivism (re-commission of an unlawful act) by the defendant and the term of imprisonment. Independent studies of sentencing established a discriminatory racial approach to sentencing algorithms: a stricter sentence for black people and a milder sentence for white people. All systems of algorithms used in judicial practice in the USA are patented and protected by the law on industrial models and trademarks. Therefore, members of the judicial community do not have access to the programs embedded in the algorithm.

In 2018, the Council of Europe adopted the European Ethical Charter on the use of Artificial Intelligence in judicial systems and the surrounding realities, which emphasizes the danger of discrimination in the judicial process and in decision-making, since the algorithm identifies specific factors of public threat and danger in relation to ethnic religious and socio-economic factors. Thus, there is a real danger of unjustified tightening of punishment by a robot judge.

However, are robotic judges with algorithms embedded so cruel and bloodthirsty?

In 2006, the Chinese press reported on law enforcement practice in one of the provincial towns, in which a high-complexity computer issued fifteen hundred criminal sentences.

Observers were struck not only by the high performance of the machine, but also by its humanity: half of the cases examined in court provided for capital punishment, that is, execution. However, the algorithm did not pass a single death sentence! The name of the book by Friedrich Nietzsche “Human, all too human” comes to mind...