Generative Artificial Intelligence (GenAI) has emerged as a ransformative and disruptive force in education and society, with the potential to both create and correct misinformation. In education, misinformation manifests at three levels: the individual (when students hold misconceptions or inaccurate beliefs); the community (when groups of individuals share the same misconceptions or inaccurate beliefs); and the system (when educational policies and practices are not based on scientific evidence). We conducted a systematic scoping review to identify existing challenges and opportunities for misinformation generation, sharing, and correction across these levels in education, as well as inform future research in this area of work. Our results indicate three approaches to GenAI and misinformation in education: identifying misconceptions, prebunking, and (re)producing misinformation, with most research conducted at the individual level. We conclude with suggestions for future research and practice.