The emergence of Generative Artificial Intelligence (GenAI) creates new challenges and opportunities within Higher Education (HE), most notably around pedagogy, academic integrity, and ethical practice. For this project, several factors determined our choice of topic:

• The public release of free ChatGPT 5 in November 2022, which caused widespread disruption across HE in Ireland and worldwide.
• The resulting uncertainty in HE about the potential impact of GenAI technologies, particularly around academic integrity, and assessment practices.
• The erstwhile lack of a GenAI policy or robust guidelines in In general, the HE sector seemed unprepared for the GenAI storm.
• The scarcity of empirical research or factual evidence, beyond speculation and anecdotes, regarding students’ actual behaviours and perceptions of GenAI, as well as their training needs.

The literature review conducted for this project highlighted several trends in relation to GenAI in education. First, it showed that students are using GenAI tools, and while relatively sparse, there is growing empirical research investigating their behaviours and usage, although the student voice is frequently absent from this research. Second, research revealed that while students generally perceive GenAI positively, they also have a strong desire for clear pedagogical and ethical guidance and training from their institutions.

Our overall goal was to partner with students in UCD’s College of Social Sciences and Law (CoSSL) to develop in-depth knowledge about the impact of Generative AI, specifically Large Language Models (LLMs), on teaching, learning, and assessment.

Three key objectives formed the basis for the project:

• To discover students’ knowledge, perceptions, and training needs in relation to GenAI/LLMs in UCD’s CoSSL.
• To discover if and how students in CoSSL are currently using GenAI/LLMs for academic work.
• To prepare students in CoSSL to use GenAI/LLMs responsibly and in line with academic integrity.

We anticipated a range of outcomes following the completion of the project, namely:

1. Evidence-based insight into CoSSL students’ current knowledge of GenAI/ LLM tools, their perceptions, and their usage behaviours in relation to academic work.
2. Evidence-based insight into the expressed training needs of CoSSL students relating to GenAI/LLMs.
3. A student-led college-wide information campaign to increase knowledge and awareness of GenAI/LLMs among students.
4. Bespoke, student-designed training resources on using GenAI/LLM tools ethically and responsibly in university work.
5. An important contribution to the existing discourse around GenAI/LLMs through the production and dissemination of multiple research outputs, including peer-reviewed papers, conference presentations and/or posters.

To complete our team, we recruited six student partners: two doctoral students, two graduate students and two undergraduate students.

1. In October 2023, we carried out a 30-item online survey of undergraduate and graduate students across CoSSL to explore their perceptions, behaviours, and training needs around GenAI/LLMs, achieving 1553 Student partners contributed both to the design of the survey questionnaire and the analysis of the survey data.
2. To explore survey findings in greater depth, three focus groups were held with students in November 2023. The focus groups were run by the undergraduate and graduate student partners, who recruited participants, organised consent forms, moderated and audio-recorded the sessions, and created session transcripts for analysis.
3. The awareness campaign, which included the creation of a project website, posters, leaflets and a podcast, was managed by the student partners, who developed content, chose the design of the different artefacts, moderated and edited the podcast, and arranged the distribution of the posters and leaflets across CoSSL.
4. Three interactive online tutorials to support foundational GenAI training were also produced by the team. The tutorial content, including voiceover scripts, was created by the graduate student partners, who also liaised with educational technologist Andrew Woods to develop the learning objects using Articulate software.

Key Deliverables and Outcomes

Insights for Teaching: The project uncovered valuable insights into student perceptions and behaviours concerning Generative AI (GenAI). Data gathered through surveys and focus groups highlighted key areas for discussion in class, shaping a more student-centred approach to GenAI training. This ensures that teaching strategies are better aligned with students' understanding and usage of GenAI tools.

Awareness Campaign: The poster, leaflet and podcast campaign launched in April 2024 successfully raised awareness of GenAI and its impact on academic learning within the College of Social Sciences and Law (CoSSL). This initiative promoted engagement with the topic, fostering an informed academic community.

Online Tutorials: Three interactive online tutorials on GenAI were developed, offering immediate value to CoSSL staff. These tutorials, available as standalone learning activities or SCORM objects for integration into Brightspace, provide foundational training on GenAI without requiring instructors to possess expert knowledge. The tutorials are licensed under Creative Commons, allowing for broad sharing beyond the university. This ensures that the materials can benefit a wider academic audience and contribute to global discourse on GenAI education.

Skill Development: Student partners gained essential qualitative and quantitative research skills, project management capabilities, and communication proficiency. These skills were honed through their direct involvement in survey design, focus group facilitation, and awareness campaigns, providing them with hands-on experience in collaborative academic research.

Long-term Impact

Sustainable Training Resources: The online tutorials developed during the project will remain accessible to UCD staff, offering a long-term solution for integrating foundational GenAI training into modules. In a rapidly evolving GenAI landscape, these resources ensure instructors can provide relevant education without requiring advanced technical expertise.

Reusable Educational Tools: Posters and leaflets from the awareness campaign can be repurposed for classroom discussions, design projects, or as standalone educational resources, extending their utility beyond the initial campaign.

Future Research Directions: The typology of student uses of large language models (LLMs) identified during the project serves as a foundation for future research on GenAI in higher education. This typology provides a framework for examining the ethical and pedagogical implications of GenAI usage.

Academic Contributions: The project has contributed significantly to the academic discourse on GenAI, evidenced by the publication and presentation of research findings. Key outputs include:

• Publications in peer-reviewed journals (e.g., Library Trends, Big Data and Society).
• Conference presentations at international forums (e.g., LIDA 2025, CILIP Ireland/LAI Annual Conference).
• Ongoing collaborative papers, such as those in preparation for Technology, Pedagogy & Education.

These outputs reinforce the project’s role in shaping scholarly and practical understanding of GenAI in educational contexts.

Resources

The principal outputs from the project comprise educational resources that can be shared with HE instructors in UCD and beyond, as they are designated as Open Educational Resources (OERs) and licensed under Creative Commons (CC BY-NC-SA).

Online tutorials: The three online tutorials designed by the student partners and created using Articulate software may be accessed at the following links:

• An Overview of Generative AI and Large Language Models
  https://overview-gen-ai-llms.s3-eu-west-1.amazonaws.com/story.html
• Academic Challenges of GenAI Use
  http://academic-challenges-of-gen-ai-use.s3-eu-west-1.amazonaws.com/story.html
• GenAI Use and Policies at UCD
  http://genai-use-policies-at-ucd.s3-eu-west-1.amazonaws.com/story.html

The tutorials may also be embedded directly into Brightspace as SCORM objects and seamlessly integrated with the grading system.

Podcast: A podcast created for the project, Let’s Chat GPT, may be accessed here: is available on Soundcloud. https://soundcloud.com/molly-newell-232623905/lets-chatgbt-episode-1.

Publications: Other outputs from the project include peer-reviewed journal articles and conference contributions that could assist others seeking more information:

• Havelka, S., McGuiness, C. & Kerrigan, P. (2025). Students’ Perceptions and Uses of ChatGPT: Implications for Teaching Artificial Intelligence Library Trends, 73.3-4 (Publishing in May).
• Jacob, , Kerrigan, P. & Bastos, M. (2024). The Chat-Chamber Effect: Trusting the AI Hallucination. Big Data and Society. Available at: https://openaccess.city.ac.uk/id/eprint/34093/1/The%20Chat-Chamber%20Effect.pdf
• McGuinness C, Havelka S, Kerrigan P. (2024). Students' Perceptions and Uses of ChatGPT: Implications for Educators. Paper presented at CILIP Ireland/LAI Joint Annual Conference, Newry, April 24th, 2024.