Using TEQSA AI toolkit resources well means treating them as inputs to four institutional outputs: a credible action plan owned by governance, an audit of every assessment against the risk generative AI poses to it, a revised set of academic integrity and assessment policies, and a standing report to the academic board that shows the plan being executed. The resources themselves are advice. The four outputs are what an assessor will ask to see.
This article sets out that method step by step, drawing on fifteen years of TEQSA registration and compliance work with private providers and on the sequence of AI-related publications TEQSA has issued since 2023.
What the resources are and what they are not
TEQSA's generative AI material sits in its gen-AI knowledge hub. The two anchor documents are the 2023 paper on assessment reform for the age of artificial intelligence, which set out two principles and a set of propositions favouring process-focused, program-level assessment design with secured points, and the 2025 follow-up, Enacting assessment reform in a time of artificial intelligence, which moves from principle to practice. Between them sits the June 2024 request for information, which required every provider to lodge an institutional action plan by 3 July 2024.
None of this is a Threshold Standard. The standards that apply are the ones that always applied: 1.4 on learning outcomes and assessment, 5.2 on academic integrity, and 6.3 on academic governance. The resources tell you how TEQSA is reading those standards in the light of generative AI, which is exactly why using TEQSA AI toolkit material is worth the effort. Our overview of the toolkit's contents covers each document in turn; this article is about what to do with them.
Step one: rebuild the action plan as a governance document
Most providers lodged an action plan in July 2024 under time pressure. In my experience many of those plans were written by one person, approved by email, and never revisited. TEQSA said at the time that it would follow up plans it found insufficient and consider its regulatory tools where a plan was absent, and its published next steps confirmed that the plans would inform its ongoing regulatory approach.
The first task is to take that plan back to the academic board and have it adopted as a governance document with an owner, a review date and a reporting line. The plan should name the assessment audit, the policy revisions and the reporting cycle described below as its own actions, with dates. A plan the academic board owns and receives reports on is credible. A plan that lives in a compliance officer's folder is not.
Step two: audit every assessment against the propositions
The 2023 paper's propositions give you a rubric. For every assessment in every unit, record what it assesses, whether it is process-focused or product-focused, whether it is a secured assessment point (invigilated, oral, practical, observed), whether it contributes to program-level assurance of a learning outcome, and what the risk is that a student could submit generative AI output undetected and still meet the criteria.
The audit produces a map of exposure by course. Its purpose is not to eliminate every product-based task but to make sure each course learning outcome is assured at least once at a secured point. TEQSA has made clear it does not regard detection software as sufficient on its own, and an audit that relies on detection as its control will not satisfy an assessor. The audit report goes to the academic board with recommended changes and becomes the first tracked output of the action plan.
Step three: revise the policy suite
The assessment audit will show where policy is silent. The academic integrity policy needs to define permitted and prohibited use of generative AI, require disclosure of use where it is permitted, and describe how allegations are investigated when the evidence is probabilistic rather than a copied passage. The assessment policy needs to require program-level mapping of learning outcomes to secured assessment points. Unit outlines need a standard statement of what is permitted in each task.
These revisions go through the normal policy approval route, which for academic policy means the academic board. Record the dates. TEQSA reads policy revision history as evidence that the institution responded to the risk, and a policy last amended in 2019 tells its own story.
Step four: make it a standing report
The last output is the one most providers omit. The academic board should receive, at least twice a year, a short report on the action plan: audit progress, assessment changes made, integrity cases involving generative AI and their outcomes, staff development delivered, and any change in TEQSA's guidance. Compliance officers increasingly ask what evidence of AI governance TEQSA wants to see, and the answer is this report and its minutes, accumulated over time.
Providers that already produce a standing academic quality report can add generative AI as a heading rather than a separate paper. The point is that the record shows the board watching the risk continuously, not once.
My view: using TEQSA AI toolkit material as a lens, not a checklist
I have watched providers try to comply with TEQSA's AI material by copying its language into a policy. Assessors see through that immediately, and since the move to self-assurance they have become alert to generic, AI-drafted compliance documents in general. The value of using TEQSA AI toolkit resources lies in applying them to your own assessments, your own policies and your own board papers, and leaving a record that shows you did. Our broader piece on AI considerations in higher education sets the wider context.
Download the AI Governance Checklist
— a one-page checklist covering the action plan, assessment audit, policy revisions and board reporting cycle, drawn from our TEQSA registration and governance work with private providers. Get the checklist
Want the full article?
Enter your email for free access to the rest of this guide and our TEQSA resource library.
Frequently asked questions
Is TEQSA's generative AI toolkit mandatory?
No. The resources are guidance and are not Threshold Standards. The obligations that apply are the existing standards on assessment, academic integrity and academic governance, which TEQSA reads in light of generative AI.
What was the 2024 request for information?
In June 2024 TEQSA required every registered provider to lodge, by 3 July 2024, a credible institutional action plan overseen by governance addressing the risk generative AI poses to award integrity, and said it would follow up insufficient or absent plans.
Does TEQSA accept AI detection software as a control?
TEQSA has not endorsed detection software as sufficient on its own. Its published approach favours program-level assessment design with secured assessment points at which learning outcomes are assured directly.
Who should own the generative AI action plan?
The academic board, as the body responsible for academic quality under Standard 6.3, with a named executive owner and a standing report on progress. A plan without governance ownership will not be read as credible.
Dr Brendan Moloney is CEO of Darlo Higher Education, Australia's largest specialist TEQSA consultancy. He holds a PhD from the University of Melbourne, is a Cambridge University Press author on governance in higher education, and has advised private providers on registration and course accreditation for more than fifteen years.