In this AI Ethics, Safety, and Responsible Use in Schools Flint Campfire Session, we focused on how schools can move beyond reactive AI rules and instead build proactive, community-driven AI ethics frameworks. The conversation centered on practical policy design, early AI literacy, data privacy, moderation systems, and how Flint supports safe, pedagogy-first AI implementation across grade levels.

Slides from the presentation can be found here.

Content covered in this session includes:

• Context-setting on school AI policy design, featuring a case study from Yokohama International School where a cross-stakeholder task force (leadership, teachers, staff, parents) created a shared AI ethics booklet to establish common expectations and language around responsible use.

• Practical rollout guidance for school AI frameworks, stressing that effective AI policy requires collaboration, visibility, and ongoing discussion rather than a one-time document.

• Embedding AI guidance directly inside Flint, including using policy documents and AI literacy materials inside chats so both students and teachers learn responsible usage within the same environment where they practice it.

• Early AI literacy for young learners, highlighting the importance of teaching students the difference between humans and machines before heavy interaction, especially for students already familiar with voice assistants.

• Relational boundaries with AI, discussing concerns about students forming overly personal relationships with chatbots and the role of proactive digital citizenship instruction.

• Moderation and student safety workflows, including flagging protocols, counselor involvement, and transparent communication processes when concerning behavior appears in student interactions.

• AI as a thought partner in the classroom, showing how assistive AI supports differentiation, accessibility (audio/oral responses), and deeper insight into student understanding rather than replacing teaching.

• Pedagogy-first AI integration principles, distinguishing assistive AI (questioning, analysis, scaffolding) from generative shortcut-seeking use, and emphasizing that instructional goals should determine AI use—not the tool itself.

• District-wide AI literacy playbook strategies, including classroom posters, shared prompting frameworks, and school-wide initiatives that normalize responsible AI habits.

• Collaborative AI learning culture, proposing events where students and teachers share AI projects to encourage low-risk experimentation and peer learning.

• Data privacy and retention discussion, clarifying that student data is not used to train models and explaining the contained environment approach to school AI safety.

• Mission & background and policy context inside Flint, demonstrating how schools can embed institutional values, policies, and branding so Sparky tailors responses to local expectations.

• Moderation customization capabilities, allowing schools to define how AI responds to harmful language, self-harm concerns, or policy violations.

• Future direction toward student profiles and automated accommodations, aiming for context-aware support across subjects while maintaining teacher oversight.

• Overall shift from restriction to responsible empowerment, reinforcing that safe AI adoption comes from education, transparency, and shared norms—not banning tools.

• Community collaboration and next steps, encouraging continued discussion, parent communication resources, and ongoing refinement of AI practices as schools learn together.

Got more questions, comments, or feedback for this topic? Feel free to raise them within the Flint Community.