Skip to main content

2026: ASCILITE 2026 Conference Proceedings

T. Cochrane, V. Narayan, T. Poole, P. Lai, P. Lake, O. Moore,  L. Li & R. Vanderburg. (Eds.), The Future is Connection: Collaboration, Creativity and Compassion in a Digital Society. Proceedings ASCILITE 2026. Brisbane, Australia. November 29 - December 2, 2026. https://2026conference.ascilite.org/

Published: 17 November 2026

ASCILITE Conference - Full Papers

  • Social displacement and relational andragogy in generative artificial intelligence-enhanced learning and teaching

    Dr Joseph Crawford, Professor Michael Cowling

    Generative artificial intelligence (GenAI) is changing how students learn and how they seek support. This conceptual paper develops social displacement theory to explain how GenAI may replace interpersonal encounters that would otherwise contribute to social learning, belonging, and access to future support. It proposes the ABCs of Human-GenAI Support Decision-Making, in which (A) achievement, (B) burdens, and (C) connections shape whether students rely on independent effort, human support, GenAI, or a combination of these resources. It then advances relational andragogy as an institutional response suited to connected adult learning in a digital society. Relational andragogy intentionally cultivates trust, compassion, mattering, psychological safety, and mutuality while preserving learner agency. Through these relational conditions, institutions can support achievement over time, respond compassionately to external burdens, strengthen students’ anchors of connection, and develop more informed human-GenAI support choices. The framework does not assume that human support is always preferable or that GenAI use is inherently harmful. Instead, we distinguish augmentative uses that expand access or prepare students for human interaction from substitutive patterns that may progressively weaken learning communities. Practical and theoretical implications are offered for educators, student support services, institutional policy, and future empirical testing.

    Read article: Social displacement and relational andragogy in generative artificial intelligence-enhanced learning and teaching 52 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Collective sensemaking of artificial intelligence in higher education learning spaces

    Dr. Amanda K Edgar, Yogita Ahuja, Billy C C Chun, Patrick O'Doherty, Kevin Anderson, Dewa Wardak, Trish McCluskey

    Higher education is experiencing rapid and multifaceted change as artificial intelligence becomes increasingly embedded within teaching and learning. The attention of research and policy has focused on students’ perspectives and securing assessment; less is known about how staff interpret what this means for the spaces where learning occurs. This qualitative case study examines the outcomes of a design-led professional learning workshop with 23 higher education professionals from diverse roles across Australasia. Workshop artifacts, including “How might we…” statements and associated memos, were analysed using reflexive thematic analysis. Five interconnected themes emerged from the analysis: Ideological positioning of AI within learning spaces; Critical design of learning spaces in the age of AI; Guiding student-centred ecologies with access, equity and inclusion; Co-intelligence in learning spaces; and Individual anxiety alongside a collective sense of opportunity. The findings suggest that participants conceptualised learning spaces as assemblages where ethical commitments, institutional values and emotional experiences are inseparable from spatial and technological design. This case study offers empirically grounded accounts of staff sensemaking and identifies key tensions and priorities for understanding learning spaces shaped by the emergence of AI.

    Read article: Collective sensemaking of artificial intelligence in higher education learning spaces 19 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Bridging the articulation gap: A multi-agent conversational AI platform for compassionate scaffolding of technical interview skills

    Owen Noel Newton Fernando, Dominic Meng Xuan Lai, Siddhanth Manoj, Chin Ann Ong, Timothy Robert Merritt, Gwendoline Quek Choon Lang

    Technical job interviews place computing students under pressure to demonstrate algorithmic and communication skills simultaneously under live observation, producing what we term the articulation gap. Existing preparation pathways provide only partial support: coding platforms ignore communication, coaching is economically inaccessible, and current AI simulators remain reactive. We present AIView, a multi-agent conversational platform that scaffolds integrated interview competencies. A LangGraph orchestrator coordinates specialised agents for question delivery and adaptive guidance alongside verbalisation monitors and an affect-aware tone layer, with assessments anchored in the SOLO taxonomy. An evaluation with 31 students showed high usability (SUS = 73.8), strong trust in AI feedback (M = 3.94), and high recommendation intent (M = 4.29). Participants valued personalised, code-aware feedback and a low-pressure learning environment. By fostering compassionate human-AI collaboration that supports creative problem solving and connection in a digital learning environment, AIView represents a step towards democratising access to career preparation through adaptive scaffolding.

    Read article: Bridging the articulation gap: A multi-agent conversational AI platform for compassionate scaffolding of technical interview skills 21 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • From AI fatigue to compassionate practice: Building student AI literacy through continuous feedback and reflective practice

    Dr Gitte Galea, Professor Michael A. Cowling

    The rapid emergence of generative artificial intelligence (AI) presents significant challenges in higher education, often leading to student dependency and educator “AI fatigue.” Despite policy calls for assessment reform, practical, evidence-based approaches to developing critical AI literacy remain limited. This paper addresses this gap by presenting a practice-based curriculum case study and reflective inquiry investigating the structural redesign of a multidisciplinary Information and Communications Technology (ICT) unit. Moving away from surveillance and AI detection, the curriculum was redesigned to decompose high-stakes assessments into a scaffolded, low-stakes formative feedback sequence. Employing a Reflective Teaching Practice (RTP) methodology grounded in Schön’s reflective practitioner framework, the inquiry evaluates this curriculum redesign through the educator’s professional reflective diary and theoretical student persona modelling. Theoretical learning archetypes were constructed to evaluate how the scaffolded structures support diverse student AI literacy profiles. Educator reflective logs highlighted strategies for managing AI dependency and inflated expectations around AI, suggesting that scaffolded AI literacy and personally relevant assessments helped mitigate cognitive offloading. Ultimately, this inquiry suggests that navigating the AI landscape requires a return to foundational, compassionate pedagogical principles, prioritising continuous feedback and structured self-reflection to support positioning AI as a meaningful learning opportunity rather than a disruptive academic burden.

    Read article: From AI fatigue to compassionate practice: Building student AI literacy through continuous feedback and reflective practice 48 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Enhancing cultural sensitivity in university artificial intelligence principles

    Xiaofan Gu, Yize Hu, Dr Lukasz Swiatek

    Universities’ education-related artificial intelligence (AI) principles are vital statements that accessibly outline the institutions’ stances towards the technology’s use in learning. Ensuring the cultural sensitivity of these principles is vital given today’s superdiversity in higher education. This paper presents insights from research examining the cultural sensitivity of these principles, using Australian universities as its case study. The research covers 42 Australian universities, with AI principle content from 28 universities analysed using a qualitative content analysis. The findings show that the idea of cultural sensitivity is generally addressed at a surface level. The paper argues that all universities not only need to provide robust, publicly accessible educational AI principles, but also actively integrate cultural sensitivity into these principles. Based on the paper’s research insights, a novel framework is provided to support this integration.

    Read article: Enhancing cultural sensitivity in university artificial intelligence principles 59 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Sophistication without clarity: Student writing versus unedited GenAI delegation

    Zixuan Guo, Kamila Misiejuk, Sonsoles López-Pernas, Mohammed Saqr, Guanliang Chen, Eduardo Oliveira

    Writing is central to how students develop and demonstrate communicative competence, yet that development is increasingly difficult to observe as GenAI enters the writing process. This paper establishes a pre-GenAI stylistic baseline of student academic writing and characterises how unedited GenAI responses to the same assessment prompts differ from it. Using stylometric and writing analytics techniques, we analysed over 118,000 words collected in 2021 from 18 higher education students in a software engineering subject. Study 1 used the seven students who completed all four tasks to establish a within-cohort reference: the size and direction of stylistic change in the same students’ writing across two semesters and two rhetorical genres. Study 2 paired each of 59 authentic submissions with a length-matched, zero-shot response to the identical prompt from three large language models. Student writing moved toward shorter words, fewer rare terms and markedly higher readability. Unedited model outputs moved in the opposite direction on each measure, gaining lexical sophistication through a narrower working vocabulary at substantial cost to readability; the human–GenAI gap in readability exceeded the students’ own shift between semesters.

    Read article: Sophistication without clarity: Student writing versus unedited GenAI delegation 53 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • The promise of Kaospilot: Human connection for programmatic AI review

    Josephine Hook, Dr. Tomáš Záhora, Matt Chen, Andrew Junor, Zoe Fitzgerald

    Generative AI has pushed universities to review curriculum at pace and scale, raising urgent questions about assessment, academic integrity and graduate capability. At Monash University, the Programmatic Assessment and AI Review (PAAIR) project required faculties to examine assessment and AI integration at program level. This paper asks how educational designers can support teaching teams to make program-level curriculum decisions while AI practice, policy and disciplinary expectations are still unsettled. We use a practice-based reflective case study methodology across the faculties of Science, Arts and Information Technology. Data sources include workshop designs, curriculum artefacts, participant feedback, program mapping outputs and facilitator reflection. Our analysis examines how Kaospilot methods, including vision backcasting, shared canvases and visible artefacts, created structured conditions for teaching teams to work with uncertainty, surface disciplinary values and make program-scale thinking visible. The cases show both the promise and constraint of adapting Kaospilot methods to faculty-wide change. The method helped create the human conditions for programmatic AI review, but its impact depended on artefacts, documentation, leadership, governance pathways and continuing educational design work. The paper argues that AI curriculum review requires structured social conditions to enact policy settings.

    Read article: The promise of Kaospilot: Human connection for programmatic AI review 20 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Feedback on feedback: Student use of AI to interpret educator feedback

    Mafruha Mowrin Hossain, Tom Goldfinch

    The increasing availability of generative artificial intelligence (GenAI) is reshaping how students engage with learning and feedback in higher education. While existing research has largely focused on AI-generated feedback and hybrid human–AI feedback approaches, less is known about how students independently use AI to interpret educator feedback. This study investigated how Computer Science students use GenAI tools to understand and act upon educator feedback. Drawing on feedback literacy theory, an exploratory mixed-methods survey was conducted with 140 undergraduate and postgraduate Computer Science students at an Australian university. Overall, 89.0% of respondents reported using AI at least occasionally to interpret educator feedback, primarily to explain unclear comments and suggest improvements. Although students generally perceived AI explanations positively, many also recognised that AI could occasionally misunderstand educator intentions. The study contributes to emerging discussions on AI-mediated feedback literacy by conceptualising generative AI as an intermediary interpretive layer between educator feedback and student action, with implications for feedback design and the development of critical AI literacy in higher education.

    Read article: Feedback on feedback: Student use of AI to interpret educator feedback 50 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • The apprenticeship layer: What assessment reform cannot secure on its own

    Carlo Iacono

    Australian higher education has responded to generative artificial intelligence with an ambitious program of assessment reform. Structural approaches, most visibly the two-lane model, seek to sustain trust in degrees by securing some assessments while opening others to AI partnership. This conceptual paper argues that the reform consensus, now sharpened by sector attention to assurance of learning and adaptive capabilities, addresses validity more directly than formation. Assessment tasks are not only measurement occasions; they can also be sites of the cognitive work through which capability develops. Drawing on cognitive apprenticeship, desirable difficulties and cognitive offloading, the paper introduces the apprenticeship layer: the stratum of entry-level cognitive work through which expertise has historically been developed and renewed. As AI is increasingly used to perform parts of this work in workplaces and universities, institutions face a risk of certifying capabilities without reliably preserving the practice through which those capabilities are formed. The paper proposes three principles for formation-centred design: friction by design, process as both learning and evidence, and a care floor of guaranteed human attention. It concludes that valid judgement and capability formation are different questions that assessment can and should be designed to answer together.

    Read article: The apprenticeship layer: What assessment reform cannot secure on its own 42 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Student perceptions on how social presence and social emotional wellbeing contribute to collaborative knowledge construction in online learning

    Thilini Irugalbandara, Jason Watson, Freya Wright-Brough, Shantha Fernando, Trina Myers

    Collaborative knowledge construction (CKC) is a central objective of collaborative online learning environments (COLEs), yet existing research has largely focused on factors associated with successful collaboration and outcomes such as engagement, participation, and satisfaction. Comparatively less attention has been given to explaining how integrated experiences of social presence and social-emotional well-being support CKC. Addressing this gap, this study investigates the socio-relational mechanisms through which these experiences contribute to CKC. The study forms part of a broader Design Science Research (DSR) project and draws on qualitative data from semi-structured interviews with 20 undergraduate students experienced in collaborative online learning. Interview transcripts were analysed using reflexive thematic analysis employing a hybrid deductive–inductive approach. The analysis identified five complementary socio-relational enabling mechanisms: Participation Confidence, Emotional Safety and Interactional Comfort, Reciprocal Responsiveness and Peer Support, Relational Trust and Collaborative Openness, and Shared Accountability and Collaborative Commitment. Collectively, these mechanisms explain how social and emotional conditions support students’ participation in CKC processes. The findings advance collaborative online learning theory by providing a mechanism-based explanation of how CKC emerges and offer a theoretical foundation for future design-oriented research and evidence-informed learning design.

  • Conversational pedagogy in practice: Exploring a unit-embedded GenAI chatbot in higher education

    Meena Jha, Stephen Colbran, Caitlin Schiavone

    Generative artificial intelligence (GenAI) chatbots are increasingly being adopted in higher education; however, limited research has examined their pedagogical integration within learning environments. This paper reports findings from an exploratory mixed-methods case study of a GenAI chatbot embedded in a postgraduate unit at an Australian university. Designed as a pedagogically embedded learning companion, the chatbot provided contextualised guidance and functioned as a first point of contact for students seeking clarification and support. Findings indicate that students perceived the chatbot as a useful and accessible learning resource that supported engagement, reflection, and self-directed learning. Three themes emerged: first-contact pedagogical support, AI-mediated scaffolding, and human-AI complementarity. Students valued the immediacy and contextual relevance of responses while recognising the continuing importance of educator guidance. These initial findings provide insights into the design of pedagogically embedded GenAI learning companions and their potential role in supporting learning within higher education.

    Read article: Conversational pedagogy in practice: Exploring a unit-embedded GenAI chatbot in higher education 47 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Classification AI vs Generative AI in grading and feedback

    Dr Manjinder Kainth

    Marking is among the most time-consuming activities in education, and artificial intelligence (AI) is widely proposed as a remedy. Attention has centred on generative large language models, yet regulators raise concerns about their bias, consistency, and explainability. This paper contrasts generative grading with per-question classification models trained only on an educator’s own marking and deployed in a human-in-the-loop workflow. The architectures are compared against criteria drawn from regulatory guidance: transparency, construct-relevant features, determinism, and specifiable training data. A narrower role for generative models is also identified, in composing and restructuring marker-approved feedback rather than in marking. Deployment analytics from one platform are then reported, covering approximately 166,000 marked questions, 39,500 essays, and six institutions: 66.2% of questions were marked in under one minute, 77.6% to 89.5% of AI suggestions were accepted, and essay models reached a quadratic weighted kappa of 0.93 for high-confidence predictions. Questionnaire responses from earlier pilots (33 markers, 30 students) are also reported: markers described marking as faster and consistent feedback as easier to provide; most students rated feedback more consistent and higher in quality. Implications for institutions selecting or building such tools are discussed, alongside the limitations of vendor-collected data.

    Read article: Classification AI vs Generative AI in grading and feedback 59 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.
  • Learning with AI to prepare for working alongside AI

    Hassan Khosravi

    Generative AI is now woven into students’ everyday learning, supporting tasks such as writing, coding, reasoning and analysis. Yet its educational value remains uncertain, because systems built to improve immediate task performance may weaken the cognitive and metacognitive processes on which durable learning depends. At the same time, there are strong calls for education to better prepare students for an AI-rich future of work and the capabilities it demands. This paper tackles this double challenge through the design of AI tools that prioritise learning over performance while developing the capabilities that matter for working with AI: creating original work, exercising sound judgement and achieving disciplinary mastery. We first outline the design framework we adapt for building such tools, in which AI learning companions are embedded within authentic learning environments. We then present RiPPLE, a co-creation platform used by more than 100,000 students, where AI companions support learners as they create learning resources, evaluate the work of peers, and engage in personalised practice, with each activity targeting one of the three capabilities. We conclude with reflections and selected empirical findings from the development and deployment of RiPPLE, drawing out lessons for designing AI tools that put learning first.

    Read article: Learning with AI to prepare for working alongside AI 16 views. Recorded article-page views through the previous day; not unique readers or PDF downloads.

Showing 14–26 of 130 articles