copyright by www.natlawreview.com
These systems are autonomous and self-learning. There is no question that will lead to many positive outcomes. However, the flip side is the question of ethics and the challenge of ensuring that machines driven by will “behave” ethically towards living things.
The classic example of ethical challenges arising from involves the self-driven car. How do we program the car in the event that a crash into a of people is imminent? Do we program the car to protect the passengers or protect the ? What about if there is only one bystander in danger while the car holds four passengers? What about a mother dog and it’s seven puppies in the road? Obviously, there are ethical dilemmas in each of these examples as well as in a plethora of other scenarios. Who is liable?
There is an enormous question of liability. If the car’s “decision” lead to the harm or death of innocent bystanders or pups, who is liable?
An system that is fueled by data can act in unpredictable ways, yet predictability is essential in legal systems . With a self-learning system that acts autonomously and independently, the liability question positively looms.
The ethical dilemma that humankind faces is to halt the development of , which might be compared to halting the industrial revolution because of the impending abuse of labor that the revolution brought upon us. The more realistic approach is to accept that advances in technology are inevitable and are generally viewed as positive developments. Thus, humankind should create a set of rules that addresses ethical dilemma issues.
How comes up with decisions is based on the work of several people such as designers, developers, financiers, and users. This distributed agency system dictates that responsibility and liability are also distributed. The issue is that traditional ethical systems address human behavior on an individual basis.
Only recently has the law established constructs to address liability that involve the behavior of responsible parties of multiple players. This includes contractual, tort and strict liability models as well as no-fault models. These systems hold all parties responsible.[…]