MSCA WP3 Research Seminar - Ethical Requirements for AI Systems

Title of talk: Ethical Requirements for AI Systems

Ethical behavior is an integral part of people’s everyday life whether they are aware of this or not. When a teacher grades exams of dozens of students aiming at consistency and transparency, she is making ethical decisions. The same happens when doctors use ethical criteria to choose which patients should be prioritized in face of limited resources, or when people are treated fairly in job interview despite their color and gender. With the speedy development of intelligent technology, systems are more and more pervasive, and machines are also expected to make ethical decisions and act ethically. However, to ensure that, system developers must be aware of ethical principles. Moreover, such principles must guide development, thus being explicitly embedded in system engineering methods. In this talk, I argue that ontologies may play a crucial role in supporting awareness, besides being essential in the development of engineering practices leading ethical and trustworthy systems. In particular, I describe our own current efforts in the development of ontology-based requirements engineering and what that entails. We will talk about promising directions and results of this research and discuss what else is to be expected as next steps.