Syntheses

• AI literacy and assessment integrity need separate workstreams
• In an AI world, schools need visible thinking, not just policing final products
• District AI implementation needs an operating model, not just a tool rollout
• AI literacy needs a mix of interaction contexts
• Persona-based AI needs purpose, boundaries, and reality checks