Students critically evaluate AI‑generated sources: they check reliability, traceability, as well as ethical and responsible use in realistic study scenarios.
- Individual and group activity
- Source verification task
- In class or online
- Evaluation of AI-generated references
- All disciplines
- Basic
- 70 min / 1-2 lessons
- 12-40 students (groups of 3-4)
- GenAI tools
- AI-generated reference lists
- External sources for verification
- Flexible classroom setting
Short description
Students investigate the reliability of AI-generated references and factual claims. Working individually and in small groups, they audit a curated list of GenAI-produced sources tied to realistic study situations (e.g. collecting literature for an assignment). By checking accuracy, traceability, relevance and ethical use, students practice critical engagement with GenAI output and reflect on their own habits of trusting or shortcutting with GenAI.
Competence domain of the Didactic Framework: Critical Engagement
By the end of this activity, students can…
- develop evaluation strategies for GenAI outputs by questioning, analysing and reflecting on AI-related content. (FLAIR Didactic Framework: LO6)
- analyze AI-generated references and factual claims for accuracy, credibility and verifiability, and recognize typical features such as hallucinated, incorrect or unverifiable entries.
- compare AI-generated sources with human-produced academic sources to identify qualitative differences. (FLAIR Didactic Framework: LO8)
- reflect on personal GenAI use, its implications for learning and the risks of over-reliance on unverified AI-generated material with focus on academic work. (FLAIR Didactic Framework: LO9)
Instructions
The teacher generates several short lists of GenAI-produced references or factual claims (some correct, some incorrect, or unverifiable). Lists are linked to realistic student tasks (e.g. finding literature for an assignment). The lists should contain a mix of accurate, inaccurate, fabricated, unverifiable, or low-quality sources to support critical evaluation.
Briefly explain the goal of the task: practicing critical engagement with AI-generated content. Students are reminded that GenAI tools often produce plausible but incorrect references.
Students receive a mini-reference list (3–5 items) and verify each entry using library databases and search engines. They note which references are accurate, inaccurate, fabricated or irrelevant and document their verification steps. Students also assess the quality of the source. Where references are authentic, students additionally evaluate the quality and credibility of the publication source (e.g. peer-reviewed journal, reputable publisher, predatory journal, unpublished manuscript).
Students work in groups of 3–4 to compare their findings and discuss recurring patterns in AI-generated outputs. Groups identify common warning signs, such as misleading formatting, hallucinated references, fabricated authors or broken links.
Groups share key observations with the class. The teacher highlights common issues and effective verification strategies, emphasizing the importance of critical engagement with AI-generated academic content.
Students write a short reflection on how they typically use GenAI for conducting information searches and literature reviews and what they plan to change after this exercise to use GenAI more responsibly and critically in the future. This can be done as homework assignment.
Assessment
Assessment may combine a short individual reflection, an annotated list of AI-generated references with verification notes, and a collaborative group summary outlining patterns identified during the audit process.
Possible challenges
- Some students may have limited experience using library databases or academic search tools.
- As GenAI tools improve over time, fabricated references or obvious hallucinations may become less common.
- Students may focus only on the risks of GenAI use without recognizing its potential value when used responsibly.
How to adress them
- Provide a short introduction, demo, or guide for using library databases and verification tools if needed.
- Regularly update examples and reference lists to reflect current GenAI capabilities and limitations.
- Encourage balanced discussion by combining critical evaluation with the understanding that GenAI tools can still be useful when used responsibly and critically.
- Factual-claim variant: Instead of AI-generated references, students verify AI-generated factual claims.
- Optional extension: Students apply the same verification process to AI-generated references for an upcoming or ongoing assignment.
- Low-AI / No-AI variant: Provide fabricated or mixed-quality references created manually by the teacher for students who prefer not to use GenAI.
- Advanced variant: Students prompt GenAI themselves to generate reference lists for their actual assignments, then conduct the audit.
- Online setting: All steps can be performed in breakout rooms, with collaborative notes on a shared document.
- Discipline-specific adaptation: Lists aligned with field-specific citation norms (APA, MLA, IEEE, legal citations, etc.).
It is recommended to use university-specific library guides on evaluating academic sources.
Chelli, M., Descamps, J., Lavoué, V., Trojani, C., Azar, M., Deckert, M., Raynier, JL., Clowez, G., Boileau, P. & Ruetsch-Chelli, C. (2024). Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis. J Med Internet Res, 26:e53164. doi: 10.2196/53164
Watson, A. P. (2024). Hallucinated Citation Analysis: Delving into Student-Submitted AI-Generated Sources at the University of Mississippi. The Serials Librarian, 85(5–6), 172–180. https://doi.org/10.1080/0361526X.2024.2433640
Using this resource
This resource is licensed under Creative Commons BY-SA 4.0 license. Suggested citation: Flair Collaboration. (2025). FLAIR Toolkit. Teaching GenAI Competencies.
