Andrew Feenberg en de toeslagenaffaire
: Een kritische reflectie op de inzet van algoritmen voor de opsporing van fraude

Translated title of the thesis: Andrew Feenberg and the Dutch childcare benefits scandal: A critical reflection on the deployment of algorithms for the detection of fraud
  • M.W.F.M. Krott

    Student thesis: Bachelor's Thesis


    The Dutch childcare benefits scandal has shown that the deployment of algorithms for the detection of fraud has a significant social impact. To facilitate the detection of fraud committed by childcare benefits applicants, the Dutch tax authorities adopted a risk classification model. The underlying algorithm, which included self-learning elements, carried out unlawful profiling based on nationality and income. This thesis examines the social impact of fraud detection algorithms by answering the following question: can the problem of automated discrimination, in particular the aforementioned unlawful profiling performed by the risk classification model, be understood more clearly on the basis of Andrew Feenbergs critical theory of technology, and if so, does this theory also offer hints towards a solution? Feenberg introduces the notion of a democratic transformation of technology, with which he provides a promising starting point for a better understanding of the aforementioned problem. In particular, Feenberg argues that technology is underdetermined in the sense that the course of its development is not fully determined by the capitalist pursuit of efficiency. This implies that technology can be transformed, provided that other values are integrated into the technical code that governs its design. To ensure that these values reflect commonly accepted ethical principles, the intervention of oppressed interests must be stimulated. More specifically, in order to transform a risk classification model of the kind set forth, and to ensure that it embodies ethical principles like non-discrimination, transparency, reliability and human autonomy, the experts responsible for its design should engage in a so-called innovative dialogue with the public, in particular with citizens from marginalized social groups affected by the model’s predictions. It is concluded that such an innovative dialogue enables the exposure and mitigation of inter-principle tensions as well as intra-principle tensions arising in the design process. This, in turn, facilitates the adequate operationalization of the aforementioned ethical principles. However, it is also concluded that a certain degree of social resistance is needed to bring about a dialogue between experts and the public.
    Date of Award31 Aug 2022
    Original languageDutch
    SupervisorEddo Evink (Supervisor), Tom Giesbers (Supervisor) & Herman Simissen (Examiner)


    • artificiële intelligentie
    • techniekfilosofie

    Cite this