A Human-Centered Approach to Ethical AI Education in Underresourced Secondary Schools

Access to AI tools is not enough: Simply providing students with AI technology doesn't ensure they'll use it responsibly. The paper emphasizes that without intentional design incorporating ethical reasoning and human support, AI initiatives in under-resourced schools risk reinforcing existing inequities rather than reducing them.

Human mentorship matters: The course's success (97.8% completion rate among nearly 180 students across 12 schools) was built on "near-peer mentorship" and discussion-based instruction alongside technology access, suggesting that human connection and guidance are critical components of effective AI education, not just supplementary.

Students can engage with complex ethical tradeoffs: Despite serving Title I schools with limited resources, students demonstrated the capacity to "apply learning and reason about ethical tradeoffs" in AI contexts, challenging assumptions that ethical AI reasoning is only for advanced learners or well-resourced institutions.

College-level rigor is achievable and valued in under-resourced contexts: The college credit-bearing course showed high engagement and was perceived as rigorous and meaningful by both students and educators, indicating that under-resourced schools and their students should not be excluded from advanced, substantive AI education.

Responsible AI education expands students' sense of possibility: Beyond technical skills, the course appeared to expand students' "perceived post-secondary trajectory," suggesting that ethical AI education can help under-resourced students envision broader futures and increased agency in their educational paths.