Integrating Ethics into AI Learning: A Socio-Technical Approach for Youth Education

As AI continues to rapidly change our world, questions and concerns about the ethical design and deployment of AI systems mount, leading to growing calls for direct attention to the ethics of AI in the design of learning experiences for youth (Garrett et al., 2020; Grosz et al., 2019). While this has often resulted in ethics ''add-ons'' to otherwise technical coursework (Saltz et al., 2019), we present a different approach: we have designed an online learning experience for high school-aged youth that directly integrates AI technical learning with attention to the ethical issues particular AI systems raise. Specifically, we have adopted a socio-technical approach to lesson design where we ground learning experiences in resonant contexts for youth that foster understating about how human and technical decisions (e.g., about what data sets are selected, or how bias is mitigated) can contribute to the ethics of how AI systems operate. To support this design goal, we formed a youth advisory group to identify resonant contexts and better understand the issues and concerns youth have about AI. For this presentation, we will discuss our design and its grounding principles, how the youth advisory group helped shape our current approach, and discuss lessons learned from our iterative design-based research. We will also discuss our efforts at creating a pro-social space for the online learning experience, challenges we've encountered, and how we've tried to address them. This work has been supported by the NSF ITEST solicitation under grant number 2241576.