Students collaboratively write with AI, critically evaluate its contributions, and refine a structured text while reflecting on responsible AI use.
- Group activity
- Co-writing task
- In class
- Structured Human-AI Co-writing
- all disciplines (best suited for academic writing fields)
- Intermediate
- 40-60 min / 1 session
- approx. 25 students (groups of 2-3)
- GenAI tool
- External sources for verification
Short description
This activity invites students to co-write an analytical brief by alternating turns with an AI chatbot, allowing them to experience real Human–AI collaboration in action. An analytical brief is a concise, evidence-based written text that presents structured analysis and reasoned conclusions on a specific issue, often used to inform decision-making or professional judgment. Through structured turn-taking, students refine arguments, verify AI contributions, and practice responsible co-authorship while producing a concise, discipline-specific brief.
Competence domain of the Didactic Framework: Collaborative Intelligence (Humans & AI)
By the end of this activity, students can…
- decide which tasks of a project they complete independently or in collaboration with GenAI and set appropriate boundaries for responsible use. (FLAIR Didactic Framework: LO21, LO22)
- formulate a clear problem statement and context for an analytical brief.
- co-write a structured, discipline-specific analytical text by alternating turns with a GenAI chatbot and contributing original content in each cycle.
- evaluate the accuracy, relevance, and credibility of AI-generated sentences by identifying contradictions, gaps, or biased reasoning.
- revise and refine the final brief by integrating citations, clarifying arguments, and producing actionable recommendations in line with course expectations.
Instructions
Form pairs or small groups (2-3 people) and ensure each student/group has access to a GenAI tool. Students begin by writing three original sentences that introduce the context and define a focused problem statement for their analytical brief. It is important that students frame the topic from the very beginning and do not leave any blank pages for the AI. Starting examples can be found under the Further Resources section.
For the next step students provide the AI with the 3 starting sentences as well as clear follow-up instructions (e.g. adding evidence, explaining a concept, introducing a counterargument, or outlining risks and limitations). Instructions should guide the AI to produce exactly three sentences.
Students and the AI take turns contributing to the text for 4–6 cycles. After each AI-generated input, students add exactly three new, non-repetitive sentences that build on the previous content and maintain coherence.
Students critically review AI contributions, correcting inaccuracies, verifying evidence, and resolving contradictions. During this phase, the instructor provides guiding questions or evaluation prompts (e.g., accuracy, evidence, bias, and coherence) to support students’ critical review without dictating content.
Students refine and edit their briefs by improving clarity, adding citations, and formulating actionable recommendations. If desired, briefs can then be compared in a shared class matrix to identify similarities, risks, and key insights across projects.
Assessment
Assessment may combine the final analytical brief, documentation of the co-writing process (e.g. turn log), peer feedback and an individual reflection. This approach evaluates both the final text and the collaboration process, focusing on writing quality and students’ ability to critically engage with AI contributions.
Grading should focus on the accuracy of concepts, the use of credible sources to verify AI-generated content, the quality of critical reflection (e.g. awareness of bias, limitations, and responsible AI use), and the clarity and coherence of the final text.
A detailed example weighting can be found under the Further Resources section.
Possible challenges
- Students may not follow the structured turn-taking (e.g. adding exactly three sentences per turn)
- Coherence across turns may be weak, leading to fragmented or repetitive texts
- AI-generated content may be inaccurate or accepted without critical review
- Students may over-rely on AI and contribute limited original input
How to adress them
- Provide clear instructions on adding exactly three new, non-repetitive sentences in each turn
- Encourage students to actively maintain coherence by building on previous contribution
- Emphasize critical evaluation and verification of AI-generated content using reliable sources
- Reinforce the importance of original student contributions and editorial control
Possible starting points for the analytical brief:
Example 1 – Business / Management: Organizations are rapidly integrating generative AI into everyday decision-making and reporting processes. While these tools promise efficiency, concerns remain regarding reliability, bias, and overreliance on AI-generated insights. This analytical brief explores how human judgment can be maintained when co-writing business briefs with GenAI systems.
Example 2 – Education: AI chatbots are increasingly used by students to support academic writing tasks. Despite their popularity, questions remain about academic integrity and students’ ability to critically evaluate AI-generated content. This brief analyzes the role of human–AI co-writing in higher education contexts.
Example 3 – Policy / Ethics: Generative AI systems are often presented as neutral and objective tools for information synthesis. However, growing evidence suggests that these systems may reproduce bias or unsupported claims. This brief investigates how structured human–AI collaboration can reduce such risks in analytical writing.
Example weighting (can be adjusted to course goals)
- Final analytical brief: 45% (Lecturer)
- Turn log / process documentation: 20% (Lecturer)
- Individual reflection: Assessing students’ ability to articulate boundary-setting, responsibility, and transparency in Human–AI collaboration.15% (Lecturer)
- Peer evaluation / collaboration: 10% (Students)
- Optional presentation / class discussion: 10% (Lecturer 5% and students 5%)
Reza, M., Thomas-Mitchell, J., Dushniku, P., Laundry, N., Williams, J. J., & Kuzminykh, A. (2025). Co-writing with AI, on human terms: Aligning research with user demands across the writing process. Proceedings of the ACM on Human-Computer Interaction. https://doi.org/10.1145/3757566.
Hwang, A. H.-C., Liao, Q. V., Blodgett, S. L., Olteanu, A., & Trischler, A. (2024). “It was 80% me, 20% AI”: Seeking authenticity in co-writing with large language models. arXiv. https://arxiv.org/abs/2411.13032
Shibani, A., Rajalakshmi, R., Mattins, F., Selvaraj, S., & Knight, S. (2023). Visual representation of co-authorship with GPT-3: Studying human–machine interaction for effective writing. In Proceedings of the 2023 Educational Data Mining Conference. Retrieved from https://educationaldatamining.org/EDM2023/proceedings/2023.EDM-long-papers.16/
Using this resource
This resource is licensed under Creative Commons BY-NC-SA 4.0 license. Suggested citation: Flair Collaboration. (2025). FLAIR Toolkit. Teaching GenAI Competencies.
