Robots in front of the court?
The development of technology leads to huge changes in our everyday life and other fields of life have to adapt these changes as well. Robotics and artificial intelligences develop in an enormous scale, new machines and programs born which didn’t exist before so we have to find new answers in the area of law to upcoming problems and bring it under regulation. If the area remain under regulated it could lead to misuse while the over regulation could be a gap of development. An important question is: May robots and artificial intelligences have rights and liability? Can a robot be sentenced?
Self-driving cars
Self-driving cars are spreading at a great pace, these days we can buy cars in which self-driving systems are equipped. Although the main benefits of driverless cars are said to be safety and predictability it can’t be proclaimed that these systems may not be dangerous and cause even fatal accidents. When a self-driving car hits another car it drives us into a corner. Who is liable for the crash? It is a dilemma because a vehicle without driver meet with an accident, a vehicle which is able to make simple decisions on its own and not a driver controlled car.
Tesla Model S is the first car equipped with self-driving system to can be bought in commercial use. The first casualty caused by self-driving cars also connected to Tesla. In may the car crashed into a turning vehicle. /Source: Autolive/
In our days maybe we don’t feel we face difficult questions when it comes to liability of driverless systems because of the moderate number of driverless cars on roads and the lack of complicated incidents despite the fact that the preponderant part of the countries don’t have laws on self-driving cars. France and the United Kingdom and some states of the United States like Nevada, Florida are exceptions as in these states we can find separated acts and the use of driverless cars are allowed. The 1949 Geneva Convention on Road Traffic and the 1968 Vienna Convention on Road Traffic both regulate the use and operation of self-driving cars moreover ordain the liability of the driver as the driver has to be ready to take over control at any time. These regulations are not satisfactory moreover not viable as the future plans are about cars without controller parts like steering-wheel.
The „brain” of self-driving systems is a self-autonomous artificial intelligence. Their decisions depend on the way they are programmed and it makes easier to answer the question of liability. The laws of customer protection more detailed the laws of product liability could be the framework of a future regulation and are the framework in the above-mentioned American states. If the driverless car meet an accident we have to examine whether the defects of the car or the self-driving systems, the owner or another driver caused the accident.
But where starts and where ends the liability of developers and producers in this question?
There are examples when the police tried to fine self-driving cars. So far, they haven’t managed to do so. /Source: Funnyjokes/
According to laws of product liability the producer is liable for the quality of the product and for the accidental damages which were caused by the defective functioning, dysfunctioning of the product. This can happen when the product of the producer cannot put its job through. It is expected from a self-driving system to know the rules of road traffic, pay attention to other vehicles and people, weather conditions and to get around in safe but it is also expected from the producer to provide the appropriate information about the use and dangers of the product. If these terms are not realized the producers has the liability as didn’t show the appropriate scale of diligence in developing the program and informing costumers. Misrepresentation also can emerge, if the system has been sold as a self-driving system while actually it is inappropriate to be involved in road traffic without accidents. In the United Stated it triggered a huge debate who has the liability in the case of a car in which the self-driving system was added later, the producer of the car, the producer of the self-driving system or the manufacturer who build the parts in the car. In this case it is also important to determine whether the accident was caused by the defects of the car or the self-driving system or by the negligence of the manufacturer because it would be inconsequence to impeach the producer of the car or manufacturer while the self-driving system had defects or impeach the producers while the manufacturer made mistakes.
It also can happen that the customer doesn’t use the product in a proper way according to its functions- for instance the driver should check the work of the system and interfere when needed but he doesn’t pay attention- and in this case the liability is on the driver. It is interesting that the legislation of the United States- like the Geneva and Vienna Convention- stick to the concept when the driver has to be ready to interfere and take over the control as it can help to avoid crashes and during system errors while the developers like Google want to exile control elements as self-driving systems are much safer. Taking over the control has other disadvantages. These cars could provide huge help for handicapped and disabled people who are not able to interfere.
Another approach is when after the damages caused by the self-driving car the owner can be called to account. This liability conception can be incapable as doesn’t urge the producers to develop more effective systems because the liability of producers totally or partly decreased. The owner let the driverless car to participate in road traffic and with this he become liable for the car similarly to canine liability. This liability conception would rather fit completely autonomous robots which can make decisions in any part of our everyday life.
I may cross the boundaries of this topic but what would happen if the driverless car were heading toward a tunnel and a child got in front of the car? If the car swerve and hit the wall of the tunnel the passengers will die, if the car doesn’t swerve it will hit the child and definitely kill him. This is the so called tunnel problem from Jason Miller.
This problem leads us to moral questions, are anyone’s life more important to be protected by the self-driving system? Should it minimize the number of casualties or protect the passenger? According to Jason Miller the Canadian, expert of the subject, the solution is to set up a moral proxy. The passenger has the right to decide in these situations and the car act according to the person’s moral proxy. It can help to solve the problem of liability as well because the car just follow an order which is the will of the passenger. Maybe it will be the best solution taking into consideration the fact that other people don’t have the right to choose between human lives, every person could make the decision on his own about his life.
Completely autonomous robots
Another member of our team, Noémi Éltető shows us from the perspective of a psychologist how machines can learn from us and where are the current borderlines of artificial intelligence. It is clearly visible from her article too that the robots able to think substantively rather belongs to the world of science fiction. Science fiction in some way or other, autonomous robots are the future from the perspective of law but a totally different regulation will be needed. I mentioned above the liability of the owner which urge him to pay more attention to the robot. The owner knows best the potential dangers and the possible situations when the robot can hazard others. The liability of the producer is not definite, the robots of the future will be able to learn for themselves it is called deep learning which can create a nearly new entity. It would mainly depend on the surrounding what kind of stuff and things the robot learn that is why the liability would be on the owner as he can influence the mindset of the robot. This liability similar to canine liability because canines can make decisions autonomously but the owner can affect the behaviour of the animals and foresee it.
The boundaries fade away between humans and robots, both in appearance and abilities. Robotethology study the social effects of humanlike robots. On the left Nadine the robot of Singapur University can be seen who was designed to help children and elderlies. Source: RT [endif]--
The European Parliament just showed a draft according to which it would be worth to create electronic personality for robots and give rights, obligations to them moreover the owner of the robots should pay social security. More details about rights and obligations are not known yet but the culpability of robots definitely won’t come true. The aim of punishment to deter from committing crimes but in the case of a robot this would be impossible as robots don’t have emotions like regret or shame. The robots won’t able to realize that they did something wrong. The liability of the state could be a solution and the owners would pay contributions and higher taxes but it wouldn’t urge the owners to prevent their robots from breaking the law. It would make the problem more complicated if in the future we could manage to give emotions to robots, can we regard a machine with emotions and substantive thinking as a human?
The regulation is not so important but it is worthy to think about the potential solutions which help innovation but show sufficient rigor in social questions.
The solving of a problem like this have to come to pass together by involving the whole society as this change concerns all of us. Among other things ESOF is the special platform of social dialogs of this kind, where whole topics („Living in the Future”, „Biorevolution”) deal with similar questions. This article was inspired by the ESOF session called "Can we simulate the human brain?". If you are interested in more exciting questions don’t miss the chance to follow us for new posts and articles on the webpage of Science Plane team!
![endif]--![endif]--![endif]--
Sources:
