New regulations in the Medienstaatsvertrag (MStV) present the state media authorities with new, diverse tasks in the area of media intermediaries: These must disclose what happens in the background when users search for information and are obliged to make the algorithmic criteria they use to select content transparent (§ 93 MStV). In addition, they are prohibited from systematically treating journalistic and editorial content better or worse than others and thus discriminating against them (§ 94 MStV). At the present time, it is still unclear how media intermediaries are to be monitored by the state media authorities with regard to compliance with the regulations, since this requires, among other things, more detailed information about the algorithms used.
Therefore, a feasibility analysis is carried out in this project: On the one hand, it will be determined how regulatory authorities can be supported by the system in the control process. On the other hand, it will be examined which algorithms from the field of Artificial Intelligence can derive recommendations for action without full access to the algorithms used at the media intermediaries. To this end, an AI toolbox will be developed on the basis of studies with the target groups involved, which will provide examples of selected support functions with regard to the transparency requirement and on the basis of which statements on feasibility will be derived.
The research project is scheduled for three years and is funded by the Saarland State Chancellery.
Partners
Landesmedienanstalt Saarland