- Faculty/module/cohort: Environmental Law and Justice
- AI Tool(s) Used: ChatGPT, Microsoft Copilot, Padlet
- AI mode(s) used (e.g., text → image, audio → text, multimodal prompts): Text generation, image generation, discussion prompts
What was the challenge?
Students were increasingly using AI in assessments. The challenge was to help them engage critically with AI rather than rely on it uncritically.
What did you do?
Created AI-generated definitions and images for Padlet activities, asked students to critique the outputs, compared AI-generated responses from identical prompts and used the discussion to develop critical thinking before assessment.
Multimodal Use
The activity combined AI-generated images, text, Padlet discussions and classroom dialogue. Students analysed multiple AI outputs and reflected on differences and limitations.
What did this look like in practice?
Students generated images using AI, compared outputs, discussed hallucinations and inaccuracies, and used these conversations to develop their understanding of environmental law.
What was the impact?
Students engaged more actively in discussions and appeared to rely less on uncritical AI use in later assessments. AI became a catalyst for discussion rather than a replacement for learning.
Ethical or practical considerations?
Students were encouraged to recognise hallucinations, bias and limitations in AI outputs and to use AI responsibly within assessment.
Reflections and Advice
- What worked well? Using AI-generated images as discussion starters encouraged deeper engagement.
- What would you refine or change next time? Continue developing an AI-aware assessment strategy and evaluate long-term effects.
- What advice would you give to colleagues considering something similar? Teach students how to question AI rather than simply use it.
Any further reflections?
Universities should prepare students to use AI responsibly instead of avoiding it.
Quick start guide for colleagues
- Estimated time required to set up: 1–2 hours
- Digital skill level required (Low / Moderate / High): Moderate
- Common pitfalls to avoid: Assuming AI outputs are always accurate and failing to encourage critical evaluation.
- Best suited for (discipline / module type / cohort): Law, Environmental Science and discussion-based seminars
