From Biased Towards Affirmative Artificial Intelligence Tools in Education

Previous research has highlighted that AI tends to contribute to bias and the marginalisation of minority groups. This paper will focus on Artificial Intelligence Tools in Education (AITED) equality and equity in the context of nation-states in the European Union, with examples from Finland. To avoid reproducing discrimination in AITED is a basic aim enforced by the law both on EU level and on national levels in EU. So far, the discussion focuses mostly on how to decrease bias and discrimination in AI, but in this paper, we aim further and introducing the idea of affirmative measures actively promoting equality and equity. The research question for this paper is: What existing methods could help us to shape more equal and affirmative AITED? We find that promoting equality and equity instead of just avoiding discrimination takes us towards affirmative rights for minorities. In this paper, we find that there is a need to engage representatives for minorities and a minority/non-discrimination expert while shaping the AITED and to create an inclusive milieu for them. We discuss the urge to use a language that specifies each minority, and we look at the difference between non-discrimination and affirmative rights for minorities. Shaping more equal AITED could also be promoted by using the WILPF, Women’s International League for Peace and Freedom, tool from Political Economic Analyses and Critical Race Theory. Introducing datasheets that accompany every data set seems also beneficial for more equal AITED.