Universities have responded to generative AI mainly with policies, integrity warnings, and some have attempted to provide assessment guidance. While these measures are necessary, they are not sufficient. Empirical evidence from our extensive research that investigated the use of GenAI in a permitted supervised in-class assessment shows that students are calling for proper guidance. Students do not need more rules. They are calling for a systematic method.

Our empirical data, collected using our Structured AI-Guided Education (SAGE) framework at Central Queensland University, reveals a student population that is far more thoughtful than the prevailing discourse assumes. When scaffolded with structured verification requirements, 73 per cent of students verified AI outputs systematically. A further 81 per cent engaged in deep revision, either rewriting AI outputs entirely or using iterative follow-up questioning to refine them. Only 14 per cent adopted surface-level approaches such as minor formatting adjustments.

Perhaps the most striking finding concerns what students actually request from their institutions. In a recent study of 167 first-year ICT students, 51.5 per cent requested verification guidance and 33.5 per cent requested prompt examples. Combined, these figures indicate that the majority of students requested technical skills development. Students are not just asking, "What are the rules?" They are mostly asking, "How do I do this properly?"

This pattern is reinforced by the qualitative data analysed, which showed that 45 per cent of concerns centred on policy clarity, with students expressing anxiety about accidental misconduct rather than intentional dishonesty. This is a legitimation crisis, not a compliance crisis. Students have already developed AI interaction competencies through experiential learning. However, they lack institutional validation and structured guidance.

The equity dimension further strengthens this argument. The second-highest driver for AI use was English language confidence, cited by 46.7 per cent of the cohort. For students from linguistically diverse backgrounds, AI functions not as a shortcut but as an equaliser, enabling intellectual contribution without linguistic barriers. Blanket prohibitions risk removing scaffolding that levels the playing field.

Policies define boundaries, but they do not teach verification. Assessment validity can be achieved by helping students demonstrate how they engaged with AI: what they accepted, modified, or rejected, and how they made those decisions. This is an orchestration problem, and it requires an orchestration solution. SAGE is a practical orchestration framework for permitted-use assessments, developed and validated across six empirical studies involving more than 1,000 students at five Australian university campuses. The workflow comprises five steps.

In the first step, Generate, students work from a standardised prompt designed by the educator, specifying the task context, constraints, and parameters that AI cannot infer independently. This establishes a consistent baseline, ensuring equity across the cohort and models effective prompt construction through practice.

In the second, Evaluate, students verify outputs against authoritative sources. SAGE embeds this verification requirement into the assessment design itself, so that students develop cross-referencing skills through doing. If a claim cannot be verified, it cannot be used.

Third, Refine: students modify outputs with evidence-based justification, recording each decision as Accept, Modify, or Reject in a structured decision log.

Fourth, AI Critic: students submit their refined work back to AI for critique, then evaluate the evaluator, determining which feedback reflects genuine insight and which reflects algorithmic limitation.

Finally, Reflect: students produce a brief metacognitive analysis documenting what they learned about AI limitations and their own decision-making. Evidence shows that 95.2 per cent of our participants indicated they would modify their AI strategy in future assessments. This demonstrates precisely the growth mindset this phase is designed to cultivate.

SAGE does not require wholesale assessment transformation. Educators can begin with three targeted changes: a one-paragraph permitted-use statement clarifying what AI may and may not do, a lightweight decision log for two or three key outputs, and rubric criteria that assess the quality of justification, rather than the volume of AI interaction.

Generative AI is now part of the higher education environment, and it looks like it will be fully integrated within the next three to five years. The strategic question is no longer whether to permit it, but whether institutions will equip students with structured, auditable workflows that protect both validity and learning. Our students have already demonstrated, through their verification behaviours, their revision depth, and their appetite for guidance, that they are ready to be partners in this process rather than subjects requiring surveillance. Learning how to use AI ethically, critically, and efficiently is not a concession to technological convenience. It is the defining educational challenge of this decade. SAGE offers a practical, evidence-based path to meet it.

Note: The SAGE framework has been listed on TEQSA's national Generative AI Knowledge Hub. Readers can explore the framework and its resources at sage-framework.com

Blog Authors: