A Study of Latinx Youth’s Civic Engagement Through Designing Machine Learning Applications

• In Event: Empowerment-Focused Computing in Constructionist Learning Environments

Abstract

Objectives AI and machine learning systems mediate many aspects of young people’s lives—from social media to entertainment and communication. Yet, meaningful opportunities to examine how these systems work or interrogate their embedded power structures remain scarce (Morales-Navarro et al., 2023). This gap is especially concerning for Latinx students, who are often subject to stereotypes that portray them as disinterested in technology (Martínez, 2019). This study challenges such deficit framings by positioning Latinx youth as capable designers of AI applications that respond to community needs. Drawing on Papert’s (1980) constructionist vision, we explore how students engaged in critical design practices, combining locally relevant data with machine learning concepts to reimagine how such systems might operate differently to support justice. Students used a co-designed, unplugged "Creation Kit" to prototype algorithmic solutions to concerns such as food systems and labor equity.

Theoretical Framework We ground our work in constructionist and learning-through-design approaches (Kafai & Resnick, 2013), which emphasize building personally and socially meaningful artifacts. Design tasks were open-ended, collaborative, and rooted in students’ lived experiences (Shehab et al., 2025), allowing learners to define problems and iterate solutions. In line with Morales-Navarro (2025) and Iivari et al. (2020), we frame AI design as both a technical and critical process—one that requires understanding how systems work, questioning design assumptions, and reflecting on societal implications. Our definition of critical AI literacy (Long & Magerko, 2020) aligns with Nichols and Stornaiuolo’s (2019), incorporating socio-political dimensions, and our focus is on students' capacity to interrogate and reimagine AI systems in pursuit of justice.

Methodology We conducted a qualitative case study within a design-based research (DBR) framework (Brown, 1992). In partnership with teachers, we adapted “I Love Algorithms” (Carter et al., n.d.) kit to reflect local realities in a Latinx-majority community. Implemented with 10th-grade students at a California high school, the activity used a card-based prototyping tool to explore ML concepts—such as classification and clustering—in relation to real-world issues. Students designed AI applications addressing topics like wage equity, food redistribution, and surveillance, working collaboratively to articulate ethical concerns and reimagine AI toward justice. Data included audio recordings of discussions and written reflections. We drew on Vakil’s (2018) justice-oriented computing framework to analyze how students' civic and political identities emerged through design,

Results All student groups designed AI systems rooted in community concerns, revealing how civic and political identities can surface through critical design. One group critiqued surveillance bias by shifting classification algorithms toward behavioral rather than racial profiling. Another redefined “performance” in wage distribution algorithms to reflect equity values. A third group linked food waste to gentrification, designing a system to redirect surplus food to local food banks. These designs illustrated students’ emerging critical AI literacy and their capacity to reimagine systems to support collective well-being.

Significance This study shows how constructionist, design-based learning can foster critical AI literacy and help students articulate civic and political identities. Latinx youth emerged as capable designers of technologies that challenge structural injustices and envision AI as a tool for community justice.

Authors

• Santiago Ojeda-Ramirez, University of California - Irvine

• Kylie A. Peppler, University of California - Irvine

• Eva Durall, University of Oulu