What

is your opinion:

(3 points)

According to Austrian law on corporate criminal

liability, can a corporation be liable for a damage caused by wrongdoing of a

staff member ...

Imagine an AI traffic system which organises switching of traffic lights and

speed limit signs in Linz.

• Which

  four ethics principles need to be taken into account for developing the

  decision-making algorithm?

• From your point of

  view, whether and why one (which?) principle shall be decisive? Please,

  substantiate your answer!

(5 points)

A and B elaborate together an AI based system. C,

who is an intern in the company, delivers also a part of the system. A and B

take the algorithm of C without checking it of errors. The whole system causes

a damage due to a failure of Cs code. Who is causal for the damage from a

criminal law perspective?

An autonomous car (legally) drives on the street. Suddenly

there is a group of people on the on side and a child on the other side in

front of it on the street. There is no possibility to avoid killing the people

of the group or the child. The “driver” cannot take over the control of the

car. The developer A of the dilemma-algorithm coded that the system has to kill

only one person instead of a group of people in those unavoidable situations.

An autonomous car drives on the street. Suddenly

there is a group of people on the one side and a child on the other side in

front of it on the street. There is no possibility to avoid killing the people

of the group or the child. The “driver” is urged to take over the control of

the car. He decides to hit and kill the child instead of the group.

You, as an AI expert, have to develop new AI based systems.

What do you have to consider avoiding negligent acting?

What means negligence in the sense of criminal law?

A is

a developer and producer of AI-based Apps. S/he developed a new App and she paid

attention to all legal norms, all guidelines, but she forgot to evaluated all possible

risks. She also documented everything. After two years the system failed and

caused damages.

Thus,

A acted negligently.

A is a

developer of AI-based systems for autonomous cars. S/he paid attention to all legal

norms and relevant technical guidelines, and based on this evaluated all the

risks, and s/he also documented everything. However, the system failed and

caused an accident. A car passenger died.

Thus,

A acted negligently.