The rise of artificial intelligence (AI) generators, such as ChatGPT, presents both breathtaking opportunities and significant challenges for their integration into education. This study investigates the perspectives of 143 students (aged 14–16) from grammar schools (Gymnasien) across nearly all German federal states regarding the use and regulation of AI generators in school settings. Students expressed their views in essays responding to the question: “In your opinion, what limits or restrictions should there be on the use of AI generators like ChatGPT in and for school?” Their responses were analyzed using thematic coding and the Fairness, Accountability, and Transparency (FAT) framework, providing a structured understanding of their perspectives. The findings reveal distinct patterns in students’ reasoning: those advocating for unrestricted AI generator use (10%) focused primarily on accountability-related aspects, such as personal responsibility, enforcement challenges, and the future relevance of emerging technologies. Conversely, students supporting a complete ban (11%) emphasized fairness-related concerns, including academic integrity, equal opportunities, and potential classroom disruption. Meanwhile, the majority (76%) favoured restricted use, drawing on arguments from all three FAT dimensions. These findings highlight how students’ regulatory preferences align with specific ethical priorities, suggesting that their views on AI generators in education depend on the values they prioritize. This study provides valuable insights into how secondary school students perceive fairness, transparency, and accountability in AI generators’ use, offering essential guidance for policymakers and educators striving to balance technical innovation with academic integrity.