Normalizing toxicity: the role of recommender algorithms for young people's mental health and social wellbeing.

This article explores how social media recommendation systems shape the digital consumption practices of young people and the potential implications for mental health and wellbeing. It examines how the consumption of increasingly radical content, with a focus on gender-based violence and misogyny, is presented on young people's feeds in increasingly high dosages, which has significant implications for young people's social development. Employing a mixed-methods approach, this research draws on three data sources: (i) long-form interviews with young people, (ii) algorithmic analysis of over 1,000 social media videos, and (iii) roundtable discussions and interviews with school leaders from across England and Wales. These methods were used to triangulate how digital environments encourage and normalize harmful ideologies, normalizing radical content, and the affective impacts of this content on young people's wellbeing. The study presents three main findings. First, recommendation systems amplify and subsequently normalize harmful ideologies, increasing users' exposure to radical material. Second, misogynistic content is often presented as entertainment, which enables it to gain high levels of traction on social media platforms. As a result, hateful ideologies and misogynistic tropes appear in young people's behaviors, which may have significant impacts on their mental health and peer relationships. Our findings suggest the need for a significant change in approaches to digital literacy, education and policy to support young people's wellbeing and social development in digital spaces.