Professor Michael Sankey – June 2025
Recent waves of technological disruption in the education space have resulted in major changes in online learning and the how and when should students and staff work remotely. The rise of Generative AI (GenAI) has heightened significant security issues and created a widening in the digital divide between the general population, higher education, vocational education and industry. Although it is tempting to ignore the existential threat this may pose, we cannot and as these technological trends continue to evolve, it is also true that they pose significant opportunities for institutions, educators, learners and ultimately industry.
Gen AI in education and preparing students for an AI future
For the last few years now (2023/4/5), the focus was largely on the impact of GenAI on academic integrity and the need for a new, higher level of AI literacy among staff and students. However, this literacy should also encompass what the affordances are that these new technologies provide, with the focus shifting to the potential applications across the education enterprise. Notwithstanding, the national regulators across Australasia have also started to tighten the screws on institutions, requiring them to evidence how they are responding to student use of GenAI to gain an unfair advantage.
For example, in Malaysia the Ministry of Higher Education provided guidelines last year for using GenAI in higher education teaching and learning. While in Australia, the Tertiary Quality and Standards Agency(TEQSA) recently provided a strong indication that they were taking qualification integrity very seriously requiring institutions to now report on their progress in this area.
AI skills and familiarity for both staff and students is a critical issue and will form a key factor in graduate employability in the future. For institutions this means a greater emphasis being placed on discipline-based AI literacy, understanding ethical use, critical evaluations, understanding limitations and leveraging AI tools to improve productivity. While at the same time, encouraging academic integrity will continue to evolve, through curriculum and assessment reform. This is well exemplified in the Australian post-secondary sector than by the work being done by the University of Sydney and their approaches to leveraging GenAI as described by their lead in this area Professor Danny Liu.
However, there remains a very clear digital divide which poses a significant challenge for education in our region. This is clearly highlighted in the Association of Southeast Asian Nations ‘Digital Masterplan 2025’ (ASEAN, 2021) and more recently in their guidelines for AI governance and ethics (ASEAN, 2024). While in Australia, we still see nearly a quarter of people being digitally excluded, which limited access to social, educational and economic opportunities (Thomas et al., 2023). Simply put, educational providers, at all levels, need to consider how they can enable digital access, adjust their curriculum accordingly and increase student digital capabilities, particularly in the new technology, such as AI. But this is where AI can potentially help.
On a brighter note, the challenge of the digital divide is changing due to the advancements of AI technologies in everyday devices. For example, in Apple iPhones, AI (Apple Intelligence) is now there by default, benefiting some 1.3 billion users (Apple, 2024). And who can get away from the use of Copilot in Microsoft suite of tools, or Googles Gemini for those using those suites. Either way, this is a massive game changer for those who possess only basic computer skills and the support it can give them. But the vast majority of the population now have a smart device in their possession or at least access to one.
The challenges then for higher education and education more broadly is that the larger institutions, such as universities, are not renown for moving quickly, and this is often hindered due to constrained resources (most do not have an open-ended cheque book). But, with the increased need for compliance and accountability being driven by national quality agencies, we have seen delays and hurdles being experienced when looking to make changes to educational programs in response to this rapid advancement in the technology.
Academic integrity, assurance of learning and program reform
Without a doubt the number one challenge to educational institutions is that of academic integrity, which has seen a renewed focus due to waves of system-level disruptions, initially brought about by the pandemic and a rapid rise in online cheating services, but more recently due to the rise of GenAI. This has led us to a re-examination of our curriculum to ensure it can still be best fit-for-purpose to assure learning. This increased focus on integrity and assurance of learning has led to two main approaches, the first being a kneejerk reaction to bring all students back onto campus. But secondly, and more importantly, the development of new approaches to assessment that steer away from traditional online quizzes, essays and exams.
The complexity of this should not be understated, and interestingly was recognised well prior to the advent of Gen AI by Rundle et al. (2020, p. 111) who suggested that we apply a ‘Swiss cheese model’ to the assessment of core concepts, using multiple levels, or check points, across different forms of assessment. This by itself does not detract students from using Gen AI and neither it should, rather it is more about how we direct and encourage them to engage with it.
Probably the best model I have seen for this, so far, is the University of Sydney’s two lane approach to programmatic assessment. This model reflects the new reality, that teaching, learning, assessment and prepping students for their careers requires a focus on assessment for learning, where we teach students how to productively and responsibly engage with Gen AI to create, or co-create new knowledge. The trick then being, how we see this joined up across a program so that students get that meta understanding of what they have done and where it fits in the scheme of things as an emerging professional.
Moving to programmatic assessment
The concept of programmatic assessment requires a student-centred program of assessment, where multiple elements of evidence are used to build a rich picture of student performance throughout their learning journey, across all desired program learning outcomes to the given end of being a professional once leaving university.
Although a limited number of programs, such as medicine, nursing and to some extent education, already use a programmatic approach, a wider deployment of programmatic structures institution-wide is likely to take quite a few years attain. But some good example exist already, e.g. University of New South Wales has developed some great resources to support a programmatic or systemic approach to assessment. Also, Edith Cowan University is now well down the track of reforming their degree programs along systemic lines. This helpful article giving tips on their approach is most helpful. The shift to a holistic assessment of student capabilities may also see growth in transdisciplinary and interdisciplinary courses that emphasise soft and critical skills to solve complex problems, such as that being done at Australian National University.
A further strategy we see gaining a lot of momentum is peer assessments in support of programmatic assessment. This takes many forms, such as promoting early feedback to students that provide insights into how they are progressing. The aim being to support a more student-centric focus, one that emphasises partnering with students on their leaning journey. Third-party providers, such as FeedbackFruits, have also focused on peer assessments and group interactions. This approach has gained significant traction over recent years, particularly in the formative assessment space, but increasingly this is moving to new summative assessment strategies.
Further techniques such as interactive oral assessments and vivas, that are short focused conversations centred on an authentic scenarios also provide a secure and scalable method of surfacing learner understanding of the desired learning outcomes (Sotiriadou, et al., 2024). This approach can create multiple interactive touchpoints for evaluation between assessors and students, despite being mostly mediated in the online space.
The place of the voice in modern assessment
Oral assessment of course take various forms and essentially links back to the old Socratic traditions of education and today this is seen in things like objective structured clinical examination, viva voce and interactive oral assessments. These have gained significant prominence in recent years. Interactive oral assessments for example, offer opportunities to develop adaptability, problem-solving skills, industry-specific knowledge, active listening and communication skills (Sotiriadou et al.). This form of assessment has also been shown to contribute to enhancing employability skills (Colvin & Gaffey, 2023) and foster critical thinking, communication and collaboration (Tan et al., 2022). The implications of this for educational institutions are significant. For by focusing on authentic assessments like oral assessments, institutions can better prepare their students for post university life.
At the same time institutions are being asked to ensure their assessment are effectively evaluate individuals’ without being compromised by AI tools. The tendency, so far, has been to reinvent in-person examination (including oral examinations) as a way of assuring assessment of learning (Liu & Bridgeman, 2023). However, as AI continues to break new ground, the alternative is to rethink what constitutes assessment of learning to be more authentic and by virtue of that, more personalised and relevant to the students’ future work aspirations.
As mentioned earlier in this piece, this leads us back to the two-lane approach to assessment as used by the University of Sydney. Such that Lane 1 (assured assessment of learning) can be seen as authentic by including viva voces or other interactive oral assessments, and live simulation-based assessments. While Lane 2 (human-AI collaboration in assessment) can also be seen as authentic, by engaging with these new technologies in meaningful ways, thereby being authentic to contemporary discipline-based (and interdisciplinary) practices. Thus looking forward, interactive oral assessments and vivas, although coming with their own challenges for those new to this thought, show very promising results so far, especially in the context of GenAI. Not to mention that, integrating AI tools to assist us in assessment and feedback collection can also enhance both the efficiency (turnaround time) and quality (more consistent) of the process for both staff and students.
Another initiative that we see institutions adopting to support their staff is to create and support an interactive oral assessment community of practice, where staff can share ideas, assessment designs and rubrics (Ward et al., 2023), this can be done both physically and online, but mostly online now. The goal being that oral assessment practice is fulsomely embedded into the curricula to ensure that our students are equipped to embrace this new (to many) form of assessment.
ePortfolio
To facilitate the gathering of longitudinal evidence of student development and gain insight into learning processes, we are certainly seeing a resurgence in the use of ePortfolios to facilitate recording competence development, observing, reflecting across a program, and contributing to student identity formation (Sankey, 2024). As we see new forms of linking up of learning management systems, student and curriculum management systems and tools like ePortfolios, we see a new, more holistic picture of the students starting to emerge (Hicks et al., 2021). This strategic linkage of data and using the affordances of AI to provide insights around this is ushering in a new era of technology enhanced learning and changing the face of both face to face and online learning.
Renewal through educational technology as the nature of online and blended learning changes
Much attention has been paid in this piece to the fundamental shift in attention that Gen AI tools have brought over the last few years, and rightly so, as this has driven fundamental change in the ways we now approach teaching and learning. Largely because we are seeing new possibilities emerge around multimodal GenAI capability building. That is, we are finding new ways to produce text, to code, to create and manipulate images, speech, audio, video and spatial outputs. And in response we are seeing a new and vibrant diversity of outputs, one that are likely to be even more profound in the future.
At this point, AI tools are largely operate under human direction, but this may not always be the case. As soon we will see new agentic AI systems (agents) take Gen AI to new levels, based on intricate rule-based programming and tailored training, these machine learning models will move us from knowledge to action (Vohra 2024). These systems will autonomously communicate and interact with other applications to execute functions. This holds great hope for things like student support and the streamlining of many of the basic functions that are now quite time consuming in education around managing and monitoring learner records and analytics (Bijani, 2024).
More broadly, and these things run tangentially to education (inform what is to be taught), economic, social and political issues are driving demand for flexible and just-in-time learning opportunities. Global instability both financial and political has society in a place where the traditional on-campus educational experience is no longer attractive to many, being seen as too inflexible. As a result, we are seeing new approaches to teaching, learning and assessing emerge. Learners are demanding options, more flexibility, wanting engagement when ‘they’ want it, not when the institution says it will be between such-n-such hours on a particular day. Thus, as the sector grapples with meeting these demands, institutions are looking at new models of flexible and online learning that utilise AI to help them (Adabor et al., 2025).
These new forms of delivery will use a combination of in-person, online, and AI-infused activities, to allow students to benefit from interactions of a more personal nature, but drawing on extensive databases of digital resources. The clear advantage to this more flexible and accessible approach is that it enables learners from various backgrounds to access education regardless of socio-economic barriers as the technology follows them and knows where they are at. Thus it is suggested that a truly learner centred approach to education made possible through AI (University of San Diego). The potential of these new approaches lies in the ability of the systems (with human oversight) to accommodate diverse learning preferences and enhance student engagement, promoting a more inclusive educational experience.
The evolving, and rapidly changing landscape of blended and online delivery requires a significant upskilling of staff in new pedagogical approaches to effectively engage with these new learning technologies. But one must first consider that the pedagogical house needs to pull the technological cart.
Clearly, as the sector grapples with this challenge, particularly in relation to developing fair and responsive workload arrangements, the importance of this should not be understated. Rapidly changing technologies do place new, but not wholly unexpected demand on teaching staff, who are now required to monitor and respond to technological change, without necessarily having been trained in how to do so. Conceivably they will need to create AI-generated learning objects, build AI-powered agents (chatbots), use AI to help them teach and assess and to provide personalised learning experiences and individualised feedback. The trick is not to do it in isolation. The old adage remains true; it takes a village to raise a child.
Thus, as the sector faces these significant challenges, educators will develop, as they always have, new AI and technological literacies and skills, and they will develop new understandings around the ethical application of GenAI and AI more broadly. The journey has begun to reshape a future-ready educational environment, and the quest is before us. So, to finish it is appropriate to quote Frodo from The Lord of the Rings, “the road goes ever on and on down from the door where it began. Now far ahead the road has gone, and I must follow, if I can”.
Michael Sankey 