Exploring agency with and in participatory research on artificial intelligence in everyday information practices

Introduction. Artificial intelligence (AI) tools are increasingly integral to the everyday information practices of young people in digitalised societies. These tools are suggested to shape human agency; that is, human capacity to act in meaningful ways. Method. We applied participatory research methods to study how young people construct their agency in relation to AI systems as part of everyday information practices and in their participation in the research process. The data were collected with fourteen young people (fifteen-seventeen years old) and consisted of autoethnographic notes, video recordings and feedback forms. Analysis. We focused on the linguistic features of the descriptions of agency applying Jyrkämä's (2009) modalities of agency (be able, must, can, feel, want and know-how) as an analytical framework. Results and discussion. The results show that the young participants were able, could and knew how to exercise their agency in various ways both when engaging with AI technologies and when participating in a research process. However, their agency was also restricted in many ways. The modalities were interpreted to characterise the ways technologies and the research setting shape possibilities for agency and, as such, are helpful in identifying ways to develop agential practices.