Navigating ethical ambiguity without institutional guidance: generative AI decision-making among students in Zimbabwean higher education

Generative artificial intelligence tools have fundamentally altered the landscape of higher education. Students make daily decisions about appropriate AI use without meaningful institutional support. This study investigated how undergraduate learners reason about generative AI ethics in Zimbabwe, where two universities with no formal AI governance policies provided the research environment. A sequential mixed methods design combined qualitative interviews with quantitative vignette surveys across three phases. Phase one comprised semi structured interviews with 22 students across both institutions. Four critical findings emerged from integrated analysis. Five scenarios produced substantial disagreement among learners, forming genuine zones of ethical ambiguity. Students applied situational rather than universal moral reasoning to AI decisions consistently. The absence of formal institutional policy was associated with ethical norm construction shifting predominantly to peer networks, though students demonstrated sophisticated moral agency throughout. The paper argues that zones of ethical ambiguity are more plausibly attributable to institutional governance fragmentation than to student moral deficiencies. Universities bear an obligation to provide ethical frameworks for emerging educational technologies. The study contributes empirical evidence from sub-Saharan Africa and advances the institutional responsibility literature by reframing the ethical ambiguity problem as a governance concern with distributive justice implications.