Algorithms control an increasing amount of decision-making in today‘s computerized world. But where does that lead us? Instead of right and wrong, or legal and illegal, algorithms rely on probabilities. At the same time, technological developments have led to decreasing human control and oversight over them. AlgoT-research project delves into the popular concept of “algorithmic transparency” from a socio-legal perspective.
AlgoT - Potential and Boundaries of Algorithmic Transparency
The AlgoT-team includes:
Ida Koivisto, LL.D, team lead (PI). Ida Koivisto is an assistant professor of Public Law at University of Helsinki.
Riikka Koulu, LL.D, deputy PI. Riikka Koulu is an assistant professor of Law and Digitalisation at University of Helsinki, and the leader of Legal Tech Lab.
Beata Mäihäniemi, LL.D. Beata Mäihäniemi is a postdoctoral researcher at the Legal Tech Lab.
Jaakko Taipale, D.Soc.Sc. Jaakko Taipale is a postdoctoral researcher at the Legal Tech Lab.
Marta Maroni, M.Soc.Sc. Marta Maroni is a doctoral student at the Faculty of Law at University of Helsinki.
Jenni Hakkarainen, LL.M. Jenni Hakkarainen is a doctoral student at the Faculty of Law at University of Helsinki.
The AlgoT-research project focuses on “Algorithmic transparency”, the endeavor to make the inner workings of algorithmic models visible. Algorithmic transparency has been proposed as a solution to the challenges that giving algorithms a larger role in decision making can bring. Even though we often consider algorithmic models objective, there is growing evidence that algorithms can actually reproduce inequality. Algorithms may create self-fulfilling prophecies: people are no longer treated as free individuals but rather as reproducers of their previous behavior. Algorithms, therefore, have the potential to radically challenge ideas of liberalism and free will, basic tenets of western legal thinking.
Through an examination of the societal consequences of algorithmic predictions, the fundamental problem that this research project attempts to answer is as follows: what is the potential and what are the limitations of transparency in overcoming their systemic biases? This projects main hypothesis is: Algorithmic transparency fails to provide sufficient legitimacy and normative counterbalance for algorithmic decision-making. Three sub-hypotheses follow:
- against its promise, algorithmic transparency does not allow access to unmediated truth, but is performative in nature. The performance is run by legal and other experts who may be protective towards their expertise;
- the preconditions for functional algorithmic governance remain undefined;
- anthropocentric law will inevitably face a crisis in the wake of technological development.
The AlgoT research project consists of three working packages, which address each of these hypotheses. To do this, the project combines critical legal analysis with empirical qualitative research. Additionally, science and technology studies will function as uniting interdisciplinary framework. If successful, the project will provide groundbreaking knowledge on algorithmic transparency and by so doing may mitigate the societal problems caused by algorithmic logic.