Recent developments in artificial intelligence (AI) suggest we are on the brink of major transformations in the economic, political and social spheres, as well as in the domain of scientific research. These developments raise important ethical issues: is it acceptable to use recruiting or risk-prediction algorithms if they are biased against women, ethnic minorities, or any other vulnerable group? What are our obligations to be transparent, and fair, in the use of these algorithms? How do we ensure that medical robots will treat each patient with respect and dignity? What kinds of public policies will ensure the best level of preparedness for the deployment of self-driving cars, machine-learning algorithms or other AI-related technologies?
In the arena of these new and emerging issues, AI ethics is concerned with what is good, just or virtuous in the development of AI systems. This research field is both new and of a transversal nature, combining elements of the philosophy of technology, robot ethics, big data and information ethics, as well as questions of public policy, business ethics, war ethics, bioethics, political and social philosophy, and so on. More generally, AI ethics is concerned with the discovery of the moral principles that should guide the development of technologies that reproduce some of the functions of human intelligence. Initiatives such as the Montreal Declaration for the responsible development of AI aim precisely at addressing these preoccupations.
AI also raises fundamental ethical questions. For instance, is it possible to create an AI system that could be considered a moral agent? How do we create an AI system that will act for the right reasons? Should we treat some of these systems as moral patients, bearers of rights entitled to the protection of their interests? Does this suggest that these AI systems must have artificial affective or mental states? Is it even desirable to automate moral decision-making? Clearly, researchers in ethics will have a lot of work for many years to come.
For further information or to get in touch with researchers in AI ethics, please contact Martin Gibert.
Useful links :