Understanding the Shift in Australian Higher Education Assessment
The rapid integration of generative artificial intelligence into daily academic life has prompted Australian universities to fundamentally rethink how they design, deliver and assure assessments. What began as a scramble to update policies in late 2022 has evolved into a more deliberate, sector-wide effort to align pedagogy, technology and institutional governance. A recent gathering at the University of Melbourne, co-hosted with Cadmus, brought together more than 150 higher education leaders from across the country to explore these changes in depth. The conversations revealed a sector moving beyond reactive measures toward purposeful redesign that prioritises authenticity, integrity and meaningful student learning.
Australian higher education operates under the oversight of the Tertiary Education Quality and Standards Agency, known as TEQSA. This national regulator has played a central role in guiding institutions through the challenges posed by generative AI tools. TEQSA’s guidance documents emphasise that assessment must continue to provide trustworthy evidence of student achievement while also preparing graduates for workplaces where AI is commonplace. The agency’s approach recognises that detection tools alone cannot guarantee integrity and that sustainable solutions require structural changes to assessment practices.
The University of Melbourne Forum: A Catalyst for Dialogue
The University of Melbourne x Cadmus Teaching and Learning Forum served as a key moment for reflection and exchange. Moderated by Professor Jamie Evans, Pro Vice-Chancellor (Students and Education) at the University of Melbourne, the event featured panels, presentations and open discussions involving senior leaders from institutions nationwide. Participants examined how universities can design assessments that foster integrity, authenticity and human connection in an environment where AI tools are readily available to students.
Early sessions highlighted the transition many institutions have made from initial policy development and control measures to more forward-looking strategies. Leaders described a shift from “catching up” with technology to actively shaping assessment around the skills and capabilities that matter most. This evolution reflects broader recognition that assessment serves not only as a gatekeeper of standards but also as a powerful driver of learning.
Forum speakers stressed that integrity should be embedded in the design of learning experiences rather than enforced solely through surveillance. Concepts such as the Swiss Cheese Model of academic integrity were discussed, illustrating how multiple layered safeguards work together more effectively than any single approach. Critical assurance points, where supervised or secure tasks verify learning outcomes, can coexist with more flexible, creative assessments elsewhere in the curriculum.
TEQSA’s Framework for Assessment Reform
TEQSA has developed comprehensive resources to support institutions, including the 2025 publication “Enacting assessment reform in a time of artificial intelligence.” This document builds on earlier work and outlines three primary pathways that Australian providers are adopting. The guidance stresses two core principles: equipping students to engage ethically and critically with AI, and using multiple, inclusive and contextualised assessment approaches to form trustworthy judgements about learning.
The three pathways provide flexibility for institutions with different structures and contexts. Pathway one focuses on program-wide redesign, creating coherent assessment regimes aligned with degree-level learning outcomes. This approach emphasises developmental progression, shared responsibility across teaching teams and multiple secure points distributed throughout a program. It suits professional degrees with strong accreditation requirements particularly well.
Pathway two centres assurance of learning within individual units or subjects. This offers immediate visibility and addresses integrity concerns at a granular level, though it may require more secure tasks per unit. Pathway three represents a hybrid model that combines elements of both, allowing institutions to balance program coherence with unit-level safeguards where needed.
Each pathway carries distinct advantages and implementation challenges. Program-wide approaches demand significant coordination and change management but promise more efficient, valid and learning-centred systems in the long term. Unit-level assurance provides quicker wins but risks fragmentation if not carefully managed. Hybrid models offer pragmatism for complex degree structures.
Aligning Technology, Policy and Pedagogy
A recurring theme at the Melbourne forum was the necessity of coordinated action across technology platforms, institutional policies and teaching practices. Many universities are implementing large-scale reforms that respond directly to TEQSA expectations by updating digital tools alongside clear AI-use guidelines. This integrated approach avoids the pitfalls of addressing policy in isolation from the platforms that support assessment.
Examples from the sector illustrate this alignment in practice. The University of Melbourne has developed guiding principles through its Generative AI Taskforce and supports staff and students with institutionally endorsed tools. RMIT University offers a secure, private GenAI chatbot called Val for tasks such as feedback and summarisation. Other institutions are exploring oral assessments, group work platforms and in-person or supervised digital exams as ways to verify authentic student contribution.
Forum participants noted that digital platforms like Cadmus play an increasingly important role by bringing learning design, feedback and data analytics into a single environment. These tools support transparent collaboration and secure oral assessments, reinforcing the idea that technology can enhance rather than replace human connection in the learning process.
Photo by John Simmons on Unsplash
Stakeholder Perspectives Across the Sector
The Melbourne forum captured a range of viewpoints from university leaders, learning and teaching specialists and platform innovators. Pro Vice-Chancellors from institutions including UNSW Sydney, Griffith University, La Trobe University and the Australian Catholic University shared experiences of navigating change at scale. Discussions highlighted both the opportunities for renewal and the practical hurdles of implementation, such as academic workload, resistance to reduced unit autonomy and the need for extensive professional development.
Representatives from regional and smaller institutions emphasised the importance of scalable solutions that do not require disproportionate resources. Student success remained a unifying priority, with speakers underscoring that well-designed assessment supports both integrity and the development of critical AI literacy.
External experts, including Professor Phillip Dawson from Deakin University’s Centre for Research in Assessment and Digital Learning, contributed insights on the complexities of assurance in an AI-rich environment. The collective message was one of cautious optimism: meaningful progress requires ongoing dialogue and shared learning rather than isolated institutional efforts.
Challenges and Opportunities in Implementation
Despite growing consensus on the need for reform, Australian universities face several ongoing challenges. Detection of unauthorised AI use remains unreliable, shifting the focus toward design-based solutions. Resource constraints, particularly for professional development and curriculum redesign, affect the pace of change. Cultural shifts are required to move staff and students toward viewing AI as a tool for ethical, critical engagement rather than a threat to be policed.
Opportunities arise from the same pressures. Redesigning assessment around authentic professional practices can better prepare graduates for AI-enabled workplaces. Program-level approaches may ultimately reduce the overall number of high-stakes secure assessments while improving the quality of evidence of learning. Collaborative sector-wide efforts, supported by TEQSA resources and events like the Melbourne forum, help institutions learn from one another and avoid duplicating effort.
Implications for Academic Integrity Culture
The discussions at the forum reinforced that academic integrity is best understood as a cultural and capability issue rather than purely a compliance matter. Building student and staff capacity to use AI appropriately, acknowledge its limitations and reflect on its ethical dimensions strengthens the foundation for trustworthy assessment. When integrity is designed into learning activities, ethical practice becomes the default rather than an additional rule to enforce.
This cultural emphasis aligns with TEQSA’s principles, which call for assessment that equips students to participate actively in a society where generative AI is ubiquitous. It also supports broader goals of student belonging and success by making assessment feel relevant and developmental rather than punitive.
Future Outlook for Australian Higher Education
As institutions continue to implement the three pathways outlined by TEQSA, the sector is likely to see increasing diversity in assessment models tailored to disciplinary and institutional contexts. Professional accreditation bodies may accelerate the adoption of program-wide approaches in fields such as health, engineering and law. Continued investment in secure digital platforms and staff capability will be essential.
The Melbourne forum demonstrated the value of bringing leaders together to share practical insights. Similar events and ongoing collaboration through networks such as the Australian Academic Integrity Network will support consistent standards while allowing for innovation. The ultimate measure of success will be whether Australian qualifications retain their reputation for quality and integrity in a rapidly changing technological landscape.
Photo by John Simmons on Unsplash
Actionable Insights for Institutions and Practitioners
Universities can begin or accelerate reform by auditing current assessment practices against TEQSA’s principles and identifying critical assurance points. Mapping assessment across entire programs helps reveal opportunities for greater coherence and reduced reliance on high-volume detection. Investing in professional development that focuses on authentic task design and AI literacy yields benefits for both staff and students.
Platform choices should support transparency, feedback and secure options where needed. Regular review of policies, informed by sector resources and peer institutions, ensures alignment with evolving technologies and regulatory expectations. Most importantly, involving students in conversations about assessment purpose and design fosters shared ownership of integrity standards.