Reflective practice is a key component of health professions education, supporting critical thinking, professional identity development, and the integration of theory into teaching practice. At the Postgraduate Diploma in Health Professions Education at Royal College of Surgeons in Ireland, students traditionally submitted weekly reflective journals via Moodle’s Journal tool. While effective for small cohorts, this approach became increasingly unsustainable as enrolments doubled, and educators faced challenges navigating lengthy submissions, monitoring word counts, and ensuring referencing compliance.

To address these issues, we designed a scalable, human-centred reflective journalling system using Moodle’s Database activity. Each entry included structured fields for title, reflection, and references, with visible word counts. Cohorts were divided into educator-led learning sections, fostering relational engagement. Two viewing modes, list and single-entry, enabled efficient formative review while maintaining depth of feedback. Students could iteratively revise entries before final submission, supporting participatory reflection and learner agency. Complementary tools, including visual progress tracking and bespoke discussion boards, enhanced engagement and accessibility.

As generative AI becomes increasingly embedded in education, reflective activities face new challenges. Even highly personal reflections risk being replaced with AI-generated text, potentially undermining authenticity and professional growth. In response, we are exploring interventions such as AI declaration forms and multimodal reflective formats, including database-supported podcasts, to maintain ethical and meaningful engagement while scaling reflective practice.

This practitioner paper highlights a work-in-progress approach to balancing scale, human connection, and technological affordances in digital learning. It situates reflective journalling as a relational, participatory, and ethically-informed activity, designed with students and educators in mind. By sharing our design decisions, successes, and ongoing challenges, we aim to contribute to discussions about who digital learning technologies serve, how they shape engagement, and how we can preserve authentic professional reflection in the era of AI.

This practitioner paper outlines three ongoing initiatives, from the Learning, Teaching and Assessment (LTA) team at TU Dublin, responding to the ‘GenAI Teaching and Assessment Challenge’. In contrast to regulatory-focused approaches or prescriptive frameworks, these efforts emphasise agency, collaborative partnership, and developing ethical AI literacy for students and staff.

Learner Agency: The NTUTORR-funded ‘AI Driving Licence’ (AIDL) is an interactive, learner-facing resource that uses scenario-based learning to move students beyond compliance toward engaging them as partners in ethical decision-making and upholding integrity. To date, close to 1,000 learners have completed their ‘AI Driving License’, which invites them to adopt a set of 'rules of the road' designed to guide ethical decision-making in their learning and assessments.

Community practice: A SATLE Pathfinder-funded series of ‘Video Vignettes’ and podcast conversations will foreground educators’ candid reflections on redesigning assessment, and teaching practice, in an age of ubiquitous generative AI. Moving beyond polished exemplars, these peer-led case studies surface the tensions between pedagogical innovation, assessment validity, and academic integrity, offering colleagues practical reference points for their own pedagogical work.

Professional Development: A digital badge will provide TU Dublin staff with an asynchronous microlearning pathway to explore practical decision-making around generative AI and assessment. It employs a custom MS Copilot agent to scaffold, not replace, educator judgement, guide them through a decision-flow process to determine the appropriate level of AI accommodation or integration based on their specified learning outcomes, and support them in translating those decisions into clear guidance and policy statements for their learners.

Together, these ongoing initiatives contribute to an ecosystem of support at TU Dublin linking learner agency, community reflection, and professional judgment, addressing the GenAI Teaching and Assessment Challenge in ways that prioritise the human dimension and empower all involved to make ethical, evidence-informed decisions about AI in teaching and learning.

This practitioner paper discusses the use of genAI in a strategy essay-type assignment in a Business School. It explains the nature of the assignment, the preparation provided to students on the use of genAI: design of input prompts, the process of token generation using the transformer architecture of large language models, and evaluation of the output from a genAI. The paper discusses the outcome of the assignment from a learning point of view having examined the student results following the grading process. The paper compares the outcome of this work with the work of students in 2025 when students used genAI to analyse an assigned case. In 2026 an assigned case was not provided and students were allowed select an organization for analysis. The paper also considers the appropriateness of genAI (Grassini, 2023; Lappin, 2024; Lazovsky et al., 2024) for supporting strategic analysis using a variety of tools; it considers whether or not genAI provides better support for certain kinds of tool, e.g. for those tools such as PESTEL analysis that consider themes, patterns and regularities, than it does for other kinds of tools such as financial statement analysis that requires detailed analysis of numbers.

Grassini, S. 2023. Shaping the Future of Education: Exploring the Potential and Consequences of AI and ChatGPT in Educational Settings. Education Sciences, 13(7), 692. https://doi.org/10.3390/educsci13070692

Lappin, S. Assessing the Strengths and Weaknesses of Large Language Models. J of Log Lang and Inf 33, 9–20 (2024). https://doi.org/10.1007/s10849-023-09409-x

Lazovsky, Gal Sasson, Tuval Raz and Yoed Kenett. 2024. The art of creative inquiry – from question answering to prompt engineering. Journal of Creative Behavior, 59(1):1-16 DOI:10.1002/jocb.671

Journal articles, institutional/national/international guidance documents, and LinkedIn posts alike call on staff in higher education to gain, and/or improve their, ‘AI literacies’. The discourse around how to deal with the emergence of this glut of innovative, disruptive, and rapidly evolving technologies centre around: inevitability, in that the technology is already here, the students are already using it, staff should, or even must, accept these technologies and engage with these technologies. Discourse also frequently acknowledges the ethical issues accompanying different AI tools, while in the same breath saying that their use is both unavoidable and desirable. Non-engagement is is now being framed as an extreme, irresponsible position to take; individual responsibility, in that exploring and establishing ways of integrating these technologies into teaching and learning practices falls to each individual as part of their professional responsibilities; limited systemic accountability, in that the role of the institution is limited to generating policy and high level guidance, providing the typically limited degree of training and educational development in teaching and learning, and setting expectations for individual staff to develop their own capacity. As was the case before this particular ed-pocolyse, for example with the pivot to remote teaching during COVID-19, the discourse on AI technologies in higher education is frequently divorced from discussions of existing institutional digital competency frameworks and related, strategic, resourced capacity building for staff and students. Most of the ideas and proposed interventions in this space are doomed to fail as they are constructed on the shaky foundations of existing, dysfunctional dynamics in higher education. This is to say that higher education staff typically do not need to have a particular level of training or expertise in teaching and learning to be hired, and then do not have to attain any particular level when in their role, with their professional development left as a voluntary, individual endeavour. The literature is clear on higher education staff typically having overwhelming workloads, with accompanying high levels of stress and burnout, and therefore do not have a great deal of time to put into professional development, even when so motivated. Career progression pathways typically do not motivate staff to engage in time consuming professional development in the teaching and learning domain, for example, in academic roles promotion processes typically motivate staff to put more attention on the research domain. This presentation proposes an alternative, radically boring approach for how higher education staff can be positioned to be able to approach any technology, including new, innovative, and disruptive technologies, in their teaching and learning work. This approach is grounded in existing models of staff capacity building and a rejection of the nonsensical idea that higher education staff capacity is magically infinite.