Implementing

Below you will find various ways for developing critical thinking skills with AI-based tools. Examples of different activities for this purpose are grouped below by different sub-skills of critical thinking.

1. Argumentation and logical reasoning skills 

• Debate with AI: have students engage in debates with AI, challenging its answers and providing counterarguments with sources. For instance, Use AI as a “devil’s advocate” to test students’ ability to defend their viewpoint and ask them to report their discussion as an assignment to be graded.
• Evaluate AI-generated claims: present AI-generated claims (especially flawed or oversimplified ones) for students to critique and improve. Ask students to identify logical fallacies or unsupported claims in AI responses.
• Find answers to AI-generated questions: prompt AI to ask questions about the course contents to prepare for exam.

2. Analytical and synthesising skills

• Analyse concepts: ask students to define key concepts with the help of AI and to compare the results to definitions found in textbooks and research literature.
• Compare viewpoints: use AI to explore multiple perspectives on a single topic, then have students integrate those into a coherent text.
• Compare AI- and human-generated abstracts: ask students to write an abstract of an open-source research article and compare it to the AI-generated abstract.
• Write summaries: ask students to generate AI summaries or outlines and to expand the content meaningfully using concepts and references introduced during the course.
• Make critical comparisons between languages: encourage students to compare AI responses across different languages or cultural frames, and reflect on possible variations in the results.

3. Metacognition and self-reflection skills

• Document prompts: ask students to document how they used AI during a task (prompts & outputs). Discuss the results with the whole group.
• Reflect on the effect of prior knowledge: have students reflect on how their prior knowledge influenced their AI use or how AI responses confirmed/challenged their thinking.
• Compare AI and human reasoning: use AI-generated errors or limitations as a basis for reflecting on the value of human reasoning. For instance, design exercises where students compare their initial answers with AI suggestions and reflect on the differences between them.

4. Critical evaluation of sources

• Trace sources: assign students to trace claims made by AI back to original sources—or highlight when this isn’t possible.
• Reveal assumptions: explore how AI presents “neutral” information and have students analyse what perspectives or assumptions are embedded in the response.
• Check references: use fabricated references created by AI to discuss misinformation.

5. Critical AI literacy skills

• Discuss AI ethics: discuss the environmental costs and data ethics of large language models – even in cases where the teacher or the students do not want to use any AI-based tools.
• Discuss research ethics: use AI’s tendency to fabricate content as a prompt to discuss academic integrity and trust.
• Analyse biases in AI outputs: ask AI to produce answers to different questions and analyse possible biases in AI outputs (e.g., gendered translations or anglo-centric perspectives in other languages).
• Compare different AI tools: generate answers to the same question with different tools and check the answers using research literature.
• Discuss motives: explore how commercial AI tools shape what information is prioritized or excluded, and what this means for knowledge production.