GenAI & Your Thesis – Think, Check, Declare

Implementation of the Course  

The Information Literacy Hub at ETH Library developed a pilot workshop “GenAI & Your Thesis – Think, Check, Declare” as part of the mandatory doctoral lecture “Scientific Integrity for Doctoral Students” at D-ITET. It responds to the growing use of AI-based tools and generative AI in doctoral work and the need for guidance that supports reflection beyond formal rules. The workshop provides doctoral students with a protected space to reflect and discuss how generative AI affects authorship, skill development, and ethical responsibility.

Teaching Design: 
The course is delivered as a 90-minute in-person workshop embedded in a full workshop day. It is designed for a cohort of 25 doctoral candidates. In its first implementation, the course was delivered by two instructors from ETH Library. The format supports a high level of interaction and can also be delivered by a single facilitator. The teaching design prioritises active and peer learning and deliberately avoids lecture-style teaching blocks. Passive elements are limited to curated resources used during the session, while most of the workshop time is dedicated to group work and discussion.

Student engagement: 
Student engagement is structured around a sequence of interactive and collaborative activities that encourage reflection, peer learning, and transfer to the participants’ own doctoral practice.

• Opening: The session opens with a physical activity, in which participants position themselves in the room according to whether they would allow generative AI to “move in as a flatmate”. This activity serves as an entry point for discussion, making differing attitudes toward AI visible and initiating reflection on personal relationships with AI-based tools. The topic is deliberately chosen to be slightly provocative and playful, lowering the threshold for participation.
• Main: The core of the workshop is organized around a collaborative digital whiteboard (Miro1 ) with five thematic areas on the responsible use of generative AI, each providing curated resources and a shared workspace. The topics cover:
  – Literature search with AI-based tools
  – Writing with AI-based tools: style, quality, and personal consequences
  – Authorship, plagiarism, and your responsibility
  – Big brother’s watching: protecting your data
  – Declaration, transparency, and (legal) consequences
  Students form five groups, each working on one topic. Within each group, resources are distributed so that each participant focuses on a specific input and engages with the material individually, guided by predefined reflection questions. Group members then share key points and collaboratively develop “Do’s and Don’ts” for the use of generative AI related to their topic.
  The groups then present their Do’s and Don’ts to the plenary in a short pitch format. Other participants are invited to ask questions, while facilitators moderate the discussion, probe reasoning, and provide contextual clarification. This final exchange allows students to engage with the other thematic topics and compare findings across groups.
  Participants are encouraged to engage in creating creative content (AI-generated images, GIFs) visualizing and presenting their learnings to their peers, creating a high level of engagement and ownership for their own learnings.
• Closing: The session concludes with a short feedback survey inviting students to reflect on aspects they found particularly helpful and on their level of engagement. The collected feedback is used to iteratively refine and improve the course.

Pedagogical Concept 
The learning design follows a progression from personal positioning to informed peer exchange and collective articulation. By first making individual attitudes toward generative AI visible, then grounding discussion in curated resources, and finally requiring groups to formulate and publicly justify shared Do’s and Don’ts, participants move beyond opinion-based discussion towards reasoned judgment. The sharing of newly learned content enhances peer-learning, the responsibility for one thematic area per group encourages ownership and the plenary exchange supports comparison, reflection, and consolidation across perspectives.
Facilitators act as moderators and guides, supporting discussion and contextualizing student contributions rather than delivering lecture-style instructions. The course does not aim to provide technical training in AI-based tools, but focuses on responsible and transparent use aligned with ETH Zurich recommendations. It also highlights existing resources (e.g., guidelines, websites) from ETH Library and other ETH units.

Challenges and Lessons Learned: 
One challenge is addressing differing levels of prior knowledge and expectations, as the course focuses on reflection and familiarization with ETH Zurich resources and guidelines rather than advanced technical AI use. However, based on the overall positive feedback from both students and hosts, the module has been integrated permanently into the lecture and shows strong potential for scaling to other departments. One lesson learned is expectations management regarding the course objectives to lecturers and students alike, addressing the non-technical nature, reflective, and informative nature of the course.

Motivation, Project Mission, Vision Statement 

MISSION: The project aims to foster responsible and reflective research practices in times increasingly shaped by generative AI. Its mission is to equip doctoral candidates with the knowledge, critical thinking skills, and ethical awareness needed to use AI tools without compromising scientific integrity.
MOTIVATION: As the need for AI education regarding thesis work is high across institutes, departments, and teaching contexts, our intention is to offer a course that delivers the basics in a fast and flexible way. It is designed to function even in the absence of a dedicated AI expert, because it draws on curated resources (guidelines and websites available at ETH Library and other ETH units), structured discussion, and the diverse experiences of participants.
VISION: With this course, the Information Literacy Hub (ETH Library) contributes to a future-oriented learning culture at ETH Zurich, where emerging technologies are embraced thoughtfully and responsibly.

Innovative Elements 

• Integration of generative AI ethics into mandatory doctoral integrity training
• Dialogue-oriented and peer-learning-based teaching format
• Practical, case-based exercises directly linked to participants’ research practice
• Use of collaborative digital whiteboard tools to stimulate creative and critical engagement
• Focus on developing personal ethical positions rather than only transmitting rules

Effects on Student Learning 

The pilot demonstrates increased awareness of ethical AI use, improved ability to evaluate AI-generated content, and strengthened confidence in making responsible decisions.

• Feedback from the first 42 participants highlighted the following aspects:
• High perceived relevance of AI-related integrity topics
• Strong engagement through an interactive and discussion-based format
• The formulation of three Do’s and Don’ts was perceived as particularly helpful
• A “fun factor” in creatively discussing and developing one’s own stance

Which Elements of Your Project Would You Recommend to Others? 

• This course format is easily transferable and scalable to other departments and educational levels. We can train you on how to best integrate our teaching material in your teaching.
• The Miro Board is an engaging tool that leaves room for the creativity of students and thereby also addresses affective learning forms.
• Participants identify with their “artwork”, which includes guidelines of the ethical use of AI-powered tools for their doctoral project, and develop a sense of ownership as they co-create with peers.
• The resources provided for this course are to a large extend ETH Zurich guidelines and websites on the use of AI. They are highly relevant.
• We validated the course format, its transferability and achievement of the learning goals in a different department for master’s students.
• ETH Library offers genAI-related support services tailored to different levels of expertise. These services can be requested by lecturers, institutes, and departments across ETH.