The wicked challenge of designing accountability measures aimed at improving algorithmic accountability demands human-centered approaches. Based on one of the most common definitions of accountability as the relationship between an actor and a forum, this article presents an analytic lens in the form of actor and forum agency, through which the accountability process can be analysed. Two case studies - the Austrian Public Employment Service’s AMAS system and the EnerCoach energy accounting system, serve as examples to an analysis of accountability based on the agency of the stakeholders. Developed through the comparison of the two systems, the Algorithmic Accountability Agency Framework (A3 framework) aimed at supporting the analysis and the improvement of agency throughout the four steps of the accountability process is presented and discussed.

Algorithmic transparency presents a significant challenge to system developers and users of algorithmic systems alike. Framing the problem as a ‘wicked’ one, this study tackles the issue of transparency in the EnerCoach energy accounting tool through presenting a situated ethnography of the algorithmic system and exploring the issues and challenges of model transparency and post-hoc explainability therein.

According to different legal frameworks such as the European General Data Protection Regulation (GDPR), an end-user’s consent constitutes one of the well-known legal bases for personal data processing. In this paper, using an enactivist perspective in cognitive science, we develop a basic human-centric framework regarding digital consent. Based on this theoretical framework, we present our qualitative evaluation of the practice of gaining consent conducted by the five big tech companies, i.e. Google, Amazon, Facebook, Apple, and Microsoft (GAFAM).

Das Projekt untersuchte soziotechnische Dimensionen des AMS-Algorithmus auf Basis aktueller Forschung, konkret aus den Critical Data Studies und dem Bereich Fairness, Accountability and Transparency in Sociotechnical Systems. Mittels der Analyse interner und öffentlich publizierter Dokumente des AMS wurden die konzeptionelle, technische und soziale Umsetzung dieses Systems untersucht. Darüber hinaus wurden Vergleichsstudien über ähnliche Systeme in anderen Ländern herangezogen.

Transparenz, Verantwortung und Diskriminierung im Kontext von digitalem staatlichem Handeln

As of 2020, the Public Employment Service Austria (AMS) makes use of algorithmic profiling of job seekers to increase the efficiency of its counseling process and the effectiveness of active labor market programs. We analyze the system and its impact on job seekers.