AI is now inside every classroom in the MENA region, but can students and teachers actually trust the outputs it produces? This session translates cutting-edge research on trustworthy and responsible AI directly into K–12 practice, giving educators a clear, evidence-based framework for evaluating, using, and teaching AI tools with integrity. Drawing on PhD research in AI trustworthiness, this talk examines the five core dimensions of responsible AI: transparency, fairness, accountability, safety, and explainability and shows how each one applies to the tools already sitting in your school's LMS. Attendees will learn how to spot untrustworthy AI behaviour in classroom contexts, how to build age-appropriate "AI critical literacy" lessons for primary and secondary learners, and how to create school-wide responsible AI policies that go beyond surface-level acceptable use agreements. With 72% of students already using generative AI for assignments and fewer than 1 in 5 being taught responsible use, the cost of inaction is measurable. This session gives you the research and the ready-to-use tools to close that gap before your students trust an AI they shouldn't.