Helping Students Navigate AI with Purpose and Integrity (CPD Case Study)
Dr Paloma Gay Blasco (Department of Social Anthropology)
[email protected]
1.What motivated you to use AI in this module, and what goals or challenges were you aiming to address?
I am using AI in two teaching contexts:
SA4067 is an employability-focused module, designed to give students the ability to communicate their anthropological skills to potential employers. I co-convene it with Mattia Fumanti. Last year (2024-25), we dedicated one workshop to exploring AI, and we allowed AI use in one of three assessments. The workshop revolved around a discussion of the ethics of AI use, but also around what students hoped, expected, and feared in relation to AI, and then practical use of AI in connection with the assessment (so, trying it out together). This year (2025-26) we will refine and expand the use of AI, but since this is a S2 module we are still planning.
This semester (S1 2025-26) my STARIS intern Cosmo Billing and I set up the Social Anthropology SA AI Group, together with my colleague Teo Zidaru. All honours students were invited to join, and 12 did. We meet fortnightly to work directly with Co-Pilot and ChatGPT, engaging AI in a long-term simulation of a real-life anthropological situation (an anthropologist uses AI in the midst of a conflict between a local community, a government and an extractive corporation). We devise and analyse prompts together, assessing the responses we receive from AI in anthropological ways, and moving the simulation forward.
The aim is to test the affordances and limits of AI with the students, so they will learn a. to understand AI as a cultural product using their anthropological skills, b. to critique it, c. to use it skilfully.
At the same time, via this experiment and related work we are hoping to identify which might be the subject-specific, anthropological, AI-related skills that we may embed in other modules. These skills must be analytical, critical, and practical. We want to prepare our students to be effective contributors to public debate about AI, and to be able to use AI in reflexive, critical, ethically-aware and anthropologically relevant ways in the workplace.
2. How did you design or adapt the assessment, and how did you prepare students for using AI appropriately?
For SA4067 the overall module assessment is a Portfolio with three communicative pieces that the students choose out of a menu of 11. We allowed AI in the second piece. Students had to produce an Ethnographic Video/Audio or Country of Origin Report or Visual tool for research participants with an accompanying reflection. We allowed AI specifically for brainstorming, suggesting structures, and generating ideas for planning the work.
We prepared students by running the workshop ahead of the assessment, experimenting with prompts and assessing the results together. None of the students in the class (approximately 25) was keen on AI, and during the workshop the most vocal ones spoke strongly against it on ethical grounds. We experimented with different prompts and analysed the results and how they might or might not help their assessments. We considered how the resulting work might differ from standard work they might do without AI.
The end result was that no student used AI in the assessment where it was permitted.
For the SA AI Group, there is no assessment since this is a voluntary activity. It is a collaborative project in which we design the aims of each session with the students, encouraging them to consider as many different perspectives and approaches to AI as possible. It is strongly experimental, and the emphasis is on the co-production of the session and the resulting knowledge between students and staff.
3. What challenges did you encounter, and how did you address them?
For SA4067, the main challenge was our own lack of understanding of AI and of the pedagogic potential of working on AI in the classroom. So, it was very much a first, tentative attempt to begin the AI conversation in our Department, and it lacked sufficient direction. In the few intervening months our grasp of AI in connection with pedagogy has grown quite a bit. I employed Cosmo Billing as STARIS intern to gather examples of best practice in anthropological teaching and learning in connection with AI, set up an AI Working Group made up of 3 colleagues and Cosmo to plan carefully the inclusion of AI-related pedagogy in our curriculum, and set up the student Social Anthropology AI Group that meets fortnightly.
For the SA AI Group, the main challenge is that this is voluntary rather than assessed or for credit, and therefore it can only take place fortnightly so as not to overwhelm the students.
4. What benefits did you see for students and for your own teaching practice?
For SA4067, students were given a first opportunity to engage AI directly in their studies, and a space to think through and test AI anthropologically. We took our first steps in AI-related pedagogy, and this gave us the confidence to learn more, experiment and be decisive in incorporating the anthropological engagement with AI into our curriculum. This resulted in the SA AI Group, and in further experimentation with AI pedagogy in the Department by other colleagues, as well as us. We are also planning an online conference on AI and Anthropological Pedagogies to take place at the end of S2.
I am excited by the essential contribution that anthropological teaching and learning can make to this rather frightening new world of AI – in terms of critical engagement, identification of biases and sociocultural dynamics that shape AI, and anthropologically specific prompt engineering. We are looking forward to learning alongside our students the best ways to formalise and develop these skills.
5. How did you evaluate the usefulness of this assessment to ensure that it reflected the desirable learning outcomes?
For SA4067, this was our first use of AI in the Department, and on reflection it was too limited and did not have a clear pedagogic aim beyond starting the conversation around AI. Students were disappointed not to be able to use AI to generate images or sound for their video or audio, since this was where AI would have complemented their skills better. Next semester we will tie engagement with AI more clearly to specific, AI-related anthropological learning outcomes.
6. What would you do differently next time, and what advice would you give to colleagues?
Don’t let your fear of AI, or your dislike of it, prevent you from helping your students engage AI in critical, ethical, practical and subject-specific ways. Be prepared to learn alongside your students, and to develop collaborative methods to do so.