This session explores how educators working in health schools can use GenAI to support continuity, connection and teacher capacity, while avoiding uses that compromise privacy, student dignity, professional judgement or relational practice. Generative AI is rapidly becoming part of everyday educational work, offering new possibilities for planning, communication, differentiation, reflection and administrative support. In health school settings, where students experience disrupted attendance, anxiety, neurodivergence, chronic illness, hospitalisation or complex transitions between home, school and health services, these possibilities are especially significant. Used well, GenAI helps educators sustain engagement, personalise learning, reduce workload and strengthen connection. Used poorly, it introduces serious risks around privacy, student dignity, bias, over-reliance and the erosion of professional judgement. This presentation shares the development of a practical GenAI guidance approach within a health school context. The approach is built around three core questions: Is it private? Is it ethical? Is it valuable? The traffic-light model helps educators distinguish between safer, higher-value uses of GenAI and uses that require caution or should be avoided. Particular attention is given to the risks of entering identifiable student information into AI systems, automating sensitive professional judgements, and using AI in ways that weaken rather than strengthen relational practice. Workshop participants will have a scenario-based discussion, applying the traffic-light model to realistic health education dilemmas, such as using AI to support student transition planning, draft communication, adapt sensitive learning materials, summarise multidisciplinary information, or respond to student use of AI companions. These scenarios are designed to surface the grey areas where GenAI may appear helpful, but where privacy, dignity, clinical boundaries, cultural responsiveness or professional judgement is at risk. Drawing on current evidence that GenAI supports teaching productivity while also creating risks for learning, autonomy and privacy, this session positions AI as a support, not a substitute, for teacher expertise. Participants will leave with practical examples, discussion prompts and a simple decision-making framework for using GenAI in ways that protect students, support educators, and contribute to connected learning, empowered lives and improved wellbeing.