Generative Artificial Intelligence (GenAI) reproduces biased representations that cannot be addressed through technical solutions alone; children, therefore, require support in critically evaluating its outputs. While GenAI has the potential to become a space for children to reflect on bias, little is known about how parents perceive this possibility. Using gender bias as a paradigmatic case of a sensitive and value‑laden topic, this study investigates parents’ values, concerns, and expectations regarding GenAI-mediated reflection. We draw on accounts from seven families collected through a cultural probe study followed by semi‑structured interviews. The findings indicate that parents value the promotion of awareness, curiosity, discernment, respect, and pluralism. While they appreciate GenAI’s ability to prompt reflection, they oppose systems that encourage users to take action or change their beliefs. We contribute principles and requirements for reflection prompters addressing sensitive domains, and we surface the design tensions that such contexts inevitably entail.

Gender stereotypes contribute to gender inequity and emerge at a young age, highlighting the importance of raising awareness beginning in primary school.Reflection technology could help achieve this goal.However, current reflective technology is not aimed at children and does not address social issues.My thesis addresses this gap by combining HCI for Reflection with Feminist HCI and Design Activism.The results to date show that technology has the potential to prompt reflection.The current steps are aimed at confirming the findings of the explorations and better understanding the dynamics of reflection, gender, and stakeholders.My thesis will make a contribution by developing reflection tools and a method for evaluating them, as well as discussing how technology can be used to raise awareness of social issues among children.

In this pictorial, we describe an activity in which 26 10-year-old children used GenAI to create self-portraits by choosing one of four generated images, suggesting changes, and reflecting on their overall experience. By analyzing their choices and rejections, we investigated whether GenAI could elicit meaningful reflection on its limitations, particularly in representing diverse identities. While we saw moments of awareness, many children accepted results that were simply “good enough,” often focusing on surface-level changes. This Good Enough Effect emphasizes the risk of initial satisfaction limiting more profound engagement. Our main contribution is to show GenAI's ability to spark reflection while also emphasizing the importance of intentional scaffolding in moving beyond acceptance to a more critical understanding of AI and identity.

Abstract Storytelling is a vital tool for empowering marginalized groups, including those with intellectual disabilities. We can enhance their storytelling experiences through technology, making them more accessible and engaging. In this context, we introduce EmpowerBox, a multisensory storytelling toolkit co-designed with eight participants during a two-day workshop. By embracing multisensoriality—through light, vibration, sound, and tactile elements—the project encourages engagement in storytelling beyond conventional modes of expression. We present how participants crafted their stories and how the process sparked moments of social connection. Our findings highlight the transformative potential of assistive technologies in storytelling and underscore the broader societal benefits of fostering inclusive, creative collaboration.