When we think of artificial intelligence, what comes to mind are associations with science-fiction films in which robots have become the source of various problems. At the regulatory level, however, there are other kinds of challenges, such as concrete proposals to adapt already existing algorithms to social realities and to minimize problems with high-risk applications.
Dr. Joanna Mazur, a DELab UW analyst and assistant professor at the Faculty of Management at the University of Warsaw, talks about how artificial intelligence is regulated today, which processes are subject to regulation and whether regulations at the EU and international level are only about robots, answering Dr. Justyna Pokojska’s questions about:
- reasons why artificial intelligence needs to be regulated,
- the values that guide AI regulators and the requirements regulated by the European Commission,
- specific proposals emerging in the context of regulation, e.g. algorithms for automatic decision-making,
- controversial provisions in regulations concerning artificial intelligence,
- systems that analyse personal data and behaviour in order to treat citizens differently based on an assessment (social credit system),
- the definition and regulation of high-risk applications,
- moves to create robot rights,
- partnership interactions with «non-human» persons in the light of EU regulations,
- discriminatory actions embedded in algorithms.