The EU Commission is carrying on their plan to regulate artificial intelligence (AI), in order to establish binding rules for AI systems, with a particular attention dedicated to “high-risk AI systems”.
According to their newly released regulation, these include AI systems used as security components in critical infrastructure, as well as applications for crime detection and prediction, facial recognition or assessment of individuals – for example, for loans, jobs, or study places. AI applications that manipulate people, exploit knowledge of people’s vulnerabilities to harm them, or conduct general or warrantless surveillance are to be banned in general – unless the activities are permitted by law and are carried out by or on behalf of public authorities. Social scoring is to be prohibited as a matter of principle.
For WeNet, the new regulation is an important benchmark for research and development. In particular, aspects such as non-discrimination, privacy protection and transparency are central to the development of the WeNet platform.
WeNet partner PD Dr. Jessica Heesen (University of Tübingen) has commented on the regulation for the German Science Media Center – The article is available (in German) at the following link: https://www.sciencemediacenter.de/alle-angebote/rapid-reaction/details/news/eu-will-kuenstliche-intelligenz-regulieren/