AI, Ethics and Law
The Special Interest Group AI, ethics and law addresses three related themes: AI for ethics and law, responsible autonomous systems, and legal and ethical aspects of AI.
AI for ethics and law
This theme is about applying AI techniques to ethical or legal problems faced by humans. It addresses the development of AI-support for human ethical or legal decision making, investigation and compliance.
- Symbolic approaches include knowledge-based systems, argumentation models and multi-agent systems.
- Data-driven approaches apply machine learning to such tasks as text analytics, predicting outcomes of legal cases, predictive policing, discovering fraud or identifying citizens who are in need of support.
- Combinations of symbolic and data-driven approaches are studied, for instance, for explaining the results of ‘black-box’ machine-learning applications. In the law this is particularly important, since receiving an explanation of a legal decision is one of the fundamental rights of citizens, embodied in many legal procedures and regulatory frameworks.
Responsible autonomous systems
This theme is also about applying AI techniques to ethical or legal problems, but this time for making automated autonomous systems behave in ethically and legally responsible ways. Increasingly, computer systems with some degree of autonomy are being employed in practice. Such artificially intelligent software can do things that, when done by humans, are regulated by law. For example, self-driving cars have to obey the traffic laws, online information systems have to comply with data protection law, care robots can damage property or the health of the persons they care for, and autonomous weapons have to comply with the laws of war. This raises the problem of how autonomous systems can be designed in such a way that their behavior complies with ethical principles or the law.
Legal and ethical aspects of AI
This theme concerns application of the law and ethics to AI. It addresses such topics as privacy protection, freedom of speech, responsibility & liability, non-discrimination, legal and ethical aspects of self-driving vehicles, autonomous weapons and the laws of war, quality-control procedures for AI systems, and transparency requirements for algorithmic decision making. An important overarching question is whether existing regulatory frameworks suffice or whether innovative forms of regulation are needed.
Our researchers are active in the following research groups:
|Faculty of Science|
|Faculty of Humanities|
|Faculty of Law, Economics and Governance|
The SIG supports any kind of activity for promoting interdisciplinary research collaboration on the themes of the SIG. Activities can include seminars, workshops, invited researchers, grant writing collaborations. Twice a year a call for activity proposals will be distributed UU-wide, and a selection of them will receive financial support.