All laws of robotics

How will they legally regulate machines with artificial intelligence

The head of the board of directors of Mail.ru and the founder of Grishin Robotics Dmitry Grishin in mid -December presented the concept of the Law on robotics - according to him, the first in the world. In fact, such concepts, including much more detailedly developed, have been discussed for several years, including in the European Parliament. Medusa says that Grishin offers how they are going to regulate robots in the European Union, and what laws are applied to cars today.

Russian law on robotics

According to the founder of Grishin Robotics, Dmitry Grishin, he created the concept of law on robotics with the help of lawyers. The content of the document was written "Vedomosti". The author offers to equate robots simultaneously with animals and legal entities. Artificial intelligence has no emotions, but at the same time he, like animals, is capable of independent action. Grishin explains the analogy with the legal entity by the fact that the robot is also a “special legal structure”.

For robots, a special register should be created, similar to the Unified State Register of Legal Entities. The creators of robots can be applied to the same measures as owners of sources of increased danger, because the machine can cause harm in one form or another. Depending on the type of robot and the degree of danger that he carries for others, the responsibility will be on the owner or on the manufacturer: if the robot initially carries the danger and the owner is warned about this, the owner will be responsible, and the manufacturer will have to pay for injuries caused by non-gas robot, says Grishin in an interview with Republic.

Grishin offers four scenarios when the actions of the robot fall under criminal law:

• creation of a robot for offenses
• Disconnecting functions that block the possibility of harm to a person
• the creation of a robot that has no such block from the very beginning
• The construction of a robot without realizing that it can be used to harm a person

Grishin sent a concept to the Higher School of Economics, Moscow State University, St. Petersburg State University and other leading institutions in the hope that in the future it will become the basis for a separate law on robotics. In addition, he plans to recommend his achievements to European commissions. There was no reaction from institutions or representatives of the legislative power to Grishin’s proposal.

The concept of the legislation on robotics Grishin is based on the famous three laws of the robotics of science fiction writer Isaac Azimov: the robot cannot harm a person; The robot must obey a person; The robot should take care of its safety if this does not contradict the first two laws. Subsequently, Azimov added another, zero or fourth to them: the robot cannot harm humanity or allow humanity to be harmful to humanity.

What are the EU offered?

There is no separate legislation on robotics and responsibility for the behavior of autonomous robots in any country in the world. However, in the European Parliament there is a working group that dealt with this issue. Her work began recently, there are no final offers yet, but on the website of the European Parliament you can find an intermediate document that refers to the legislation for robotics.

It says that the creators of the robots should be guided by the same laws of Robotics Azimov. The Smart Robot is described in the document so a mechanism capable of autonomy by analyzing data from sensors or the exchange of information with the environment, and has the ability to self -study. Such robots should be registered at the European Agency for Robotics and Artificial Intelligence. The robots also include self -governing cars, the document emphasizes that legislative norms in relation to them should be developed and approved as soon as possible.

One of the most serious issues of any legislation on robotics is responsibility for making a robot. The document separately emphasizes that it is necessary to take into account the autonomy of robots in decision -making. The higher it is, the less responsibility for his actions should be carried by the owner. The longer the robot studied a person, the more responsibility should lie on the teacher, while it is necessary to distinguish the abilities acquired from a person with those whom the robot learned himself.

The question of who will be responsible for the damage caused by the most autonomous robot is proposed to be solved using mandatory insurance; That is, ultimately, the manufacturer, albeit in the form of insurance, is still responsible for the consequences. Experts also propose to create a fund or several funds (one for each category of robots) that will be replenished due to deductions from manufacturers and are spent on compensation for damage from the actions of robots.

The creation of a special charter of robotics is proposed, in which the ethical principles of the development of autonomous devices will be described. They are: robots should act in the interests of people; robots should not harm a person; The benefits that robots bring should be available to everyone, especially for medical robots. The principles are similar to the laws of Robotics in Azimov; It is assumed that any robot developer must keep them in mind when creating new devices. The initial code of robots should be open to the possibility of investigating incidents. For the same reasons, the history of the robot’s actions should be available.

In addition to the EU, they are engaged in the regulation of the use of robots at the UN, but from a different point of view. The issue of robots is under consideration by the UN Disarmament conference - many human rights organizations, including Human Rights Watch, believe that the use of autonomous military robots should be prohibited by conventions on the rules of war. So far, such robots are not used (operators control widespread drones), but it is quite possible to create a robot that can independently decide on the murder of the enemy.

Laws and robots right now

In the absence of special legislation, existing laws are applied to robots today. Lecturer at the University of Washington Ryan Kalo examined in detail how robots and legislation coexist in the USA. According to him, American law for six decades is trying to understand how exactly the situations with the participation of various mechanisms of varying degrees of autonomy need to be interpreted.

The American courts, as Kalo found out, more than once faced extremely unusual issues regarding robots: for example, they decided whether it was possible to say that robots perform music in the same sense as people. The scientist believes that the robot is a device that can collect data on what is happening around it, evaluate these data and act depending on its conclusions; It is these three qualities, according to Kalo, that distinguish robots from a laptop or any other high -tech device.

The question of whether the robot can be considered the same music performer as a person was raised in the United States in 1987. For entertainment, a restaurant in Maryland put in the hall several humanoid mechanisms that, according to the program in advance in them, played musical instruments. From the point of view of the legislation for any execution of live music in restaurants, you need to pay a separate tax, and the restaurant had to prove in court that these robots are nothing more than a mechanical musical machine.

After a long trial, the court still decided that such robots cannot be considered by the performers in the usual sense of the word-they act according to the program, cannot be mistaken or improvised, and, in fact, they are no different from a groovy newspaper that knocks on the drum. Then there were no robots capable of performing music without a given program, now their creation is quite possible, but how to interpret them - as performers or just an advanced music machine, is unclear.

Another curious business concerned the discovery of a sunken ship. According to marine law, the one who first finds the wreckage of a long -sunken vessel belongs to the exclusive right to everything that was on board, including gold and other values. In the late 1980s, the researchers discovered the Steamer "Central America", which sank in 1857 with a load of gold. It was possible to detect it with the help of an unmanned submarine, and a reasonable question arose: whether it belongs to the right to treasures. The court sided with the treasure hunters: he decided that people who saw the chatting of the ship were ruled by a submarine by a submarine, so there was no need to send divers to the place of collapse and additionally announce their rights. But what if the submarine is not only unmanned, but also autonomous (that is, is not controlled by a person) is unknown - there have been no similar cases in history yet.

Ryan Kalo concludes that in American law the robot is now interpreted as a programmable machine that performs the will of a person, and there were no cases when autonomous behavior would have versed. And at the moment, the situation is such that the creator or owner of the robot is responsible for his actions, although in fact there are many machines that are programmed to make independent decisions depending on what is happening around.

Pavel Borisov