The TEQSA generative AI toolkit is not a single document. It is the set of resources TEQSA has published since 2023 to help providers protect award integrity in an environment where students have free access to generative AI: the 2023 assessment reform principles, the June 2024 request for information that required every provider to lodge an action plan, the 2025 guide on enacting assessment reform, and the gen-AI knowledge hub that collects sector practice. None of it is a Threshold Standard, but together it tells you how TEQSA will read your provider against the standards it already has.
This article describes each part of the toolkit, who it is for, and how it connects to the Standards. It draws on our TEQSA compliance work with private providers, which since 2024 has included AI governance in almost every engagement.
The 2023 principles: assessment reform for the age of AI
The foundation is Assessment reform for the age of artificial intelligence, published in 2023 with sector experts. It rests on two principles. Assessment and learning should equip students for a society in which AI is pervasive. And trustworthy judgments about what students have learned require multiple, inclusive and contextualised approaches to assessment.
The propositions that follow favour assessing process rather than product, securing some assessment points so the provider can be confident the work is the student's own, and designing at the level of the whole program rather than unit by unit. For a course designer this document is the starting point, and it is the one I ask academic boards to read first. Our overview of the TEQSA framework and toolkit for generative AI sets the principles in their broader context.
The 2024 request for information
In June 2024 TEQSA moved from guidance to regulation. Its request for information required every registered provider to lodge, by 3 July 2024, a credible institutional action plan, overseen by its governance, addressing the risk generative AI poses to the integrity of its awards. TEQSA said it would follow up plans that were absent or insufficient and consider its regulatory tools.
That request is the part of the TEQSA generative AI toolkit with legal weight, because it was made under TEQSA's information-gathering powers and because the plans lodged are now part of each provider's compliance history. In my experience the plans that satisfied TEQSA were specific about which courses and units were exposed, what was changing and when, and which governing body owned the work. Plans that consisted of a policy statement and a subscription to detection software were not. Our guide to the TEQSA toolkit for providers explains what a credible plan contained.
The 2025 enacting guide
Enacting assessment reform in a time of artificial intelligence followed in 2025. Where the 2023 document set principles, this one moves to practice: how to design program-level assurance, how to choose and secure assessment points, and how to communicate expectations to students and staff. It is the most useful document in the toolkit for a provider that has an action plan and needs to make it real.
It is also the document that shows where the sector settled. The pattern TEQSA describes, and which providers loosely call a two-lane approach, is secured assessment at key points combined with open, AI-permitted learning elsewhere. That label is the sector's, not TEQSA's, but the design is now what assessors expect to see. Our guide to generative AI in higher education walks through it.
The knowledge hub and good practice resources
The gen-AI knowledge hub collects TEQSA's own publications alongside sector examples, webinars and links. It is the reference shelf. Nothing in it is binding, and some of it is already dated, but it is where a compliance officer can see what other providers have done and what TEQSA chose to publish as good practice.
The hub is best used for calibration. If your action plan looks nothing like anything on it, that is worth understanding before an assessor asks. Our summary of the TEQSA toolkit for generative artificial intelligence picks out the hub items I find most useful for private providers.
How the TEQSA generative AI toolkit relates to the Standards
This is the point I most want readers to take away. TEQSA has not amended the Threshold Standards for generative AI. It assesses AI risk under the standards that already exist: Standard 1.4 on assessment capable of confirming learning outcomes, Standard 5.2 on academic integrity, Standard 5.3 on monitoring and review, and Standard 6.3 on academic governance. The toolkit tells you how TEQSA reads those standards now, and that is different from telling you what the standards say.
The practical consequence is that a provider cannot comply with the toolkit. It can only comply with the Standards, using the toolkit as evidence of what the regulator will find persuasive. Our articles on TEQSA's generative AI resource toolkit and the toolkit for TEQSA-regulated providers each take one of those standards and show what the evidence looks like.
What I tell boards
Read the 2023 principles, check your 2024 action plan against the 2025 enacting guide, and put an AI integrity report on the academic board's agenda each semester. The TEQSA generative AI toolkit will keep growing, but the standards it points to are the same ones that governed your registration, and the record you keep against them is what will matter at renewal.
Download the Darlo AI Governance Checklist
— a one-page test of whether your action plan, assessment design and board reporting would satisfy an assessor, 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 the TEQSA generative AI toolkit mandatory?
No part of it is a Threshold Standard. The June 2024 request for information did require every provider to lodge an action plan, and that plan now forms part of the provider's compliance history, but the principles, enacting guide and hub are guidance on how TEQSA reads the existing Standards.
Which Threshold Standards does generative AI affect?
Mainly Standard 1.4 (assessment capable of confirming learning outcomes), Standard 5.2 (academic integrity), Standard 5.3 (monitoring, review and improvement) and Standard 6.3 (academic governance oversight of academic quality).
Does TEQSA require AI detection software?
No. TEQSA has not endorsed detection software as sufficient. Its publications favour program-level assessment design with secured assessment points where the provider can be confident the work is the student's own.
Do we need to update our 2024 AI action plan?
In my experience, yes. The plan lodged in July 2024 was a starting position, and an assessor at renewal will expect to see it revised against the 2025 enacting guide, with evidence that the academic board has monitored delivery.
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.