Playing and learning with AI: an exploratory emic perspective of how European children engage with Generative AI tools

Abstract (English)

Children's wider and eager embrace of generative artificial intelligence (GAI) has raised concerns among educators and parents who worry about the presumably unethical use of GAI for school related tasks, but also about more profound changes that the "affective aesthetic style embedded in its whole conversational approach" can decrease users' critical alertness needed to evaluate the truth of information provided by GAI (Balmer, 2023). At the opposite end, some scholars are enthusiastic about the opportunities that GAI brings for children.
To this etic, adult-centered perspective, our study opposes an emic perspective that aims to delve into how, why and with what results children actually use GAI as a stand-alone tool or embedded and sometimes hidden in other applications. Mapping and understanding children's digital practices with GAI as situated practices in various contexts of their daily lives helps us understand how they domesticate this new technology and how GAI reconfigure children's traditional playing, learning and communicating activities.
In doing this, we have used a qualitative methodology, with 15 in depth, semi-structured individual interviews of secondary school children that focused on their engagement with GAI tools. The interviews were part of a broader international research conducted under the umbrella of the EU Kids Online network.
Our analysis is tensed between the critical stance of datafication studies (e.g., Barassi, 2020, who talks about the datafication of childhood) and the studies that acknowledge children's agency in their interaction with the GAI (Livingstone & Third, 2017). While the conceptualization of children and their interaction with GAI is partially indebted to EU Kids Online theoretical model (Livingstone et al., 2015), we will discuss the results using the three-layer framework proposed by Davis (2023) reflecting on the fact that only starting from the feature layer of GAI and passing by its practice layer one can understand children's culture that emerge around GAI, with all its future implications (for ethics, politics, or even epistemology).