Multiplatform Journalism Course
AI was integrated into a 5-week video module of the course to support the adaptation of written news into short-form social video (Reels/TikTok/Shorts). Students used generative AI to assist prompt-based scripting, storyboard development, and AI-assisted video production, combined with reflective documentation through an AI Diary. The primary goal was to enhance multiplatform storytelling while building AI literacy and ethical awareness. A distinctive element was the structured 'human-in-the-loop' workflow, requiring students to iteratively prompt, critically evaluate outputs, and document editorial decisions across the production process.
Students showed strong functional dependence on ChatGPT, using it as a trigger or substitute for their own thinking, while paradoxically lacking basic understanding of how the tool works.
Some students interacted with ChatGPT as if it were a person or 'friend,' leading to over-reliance and uncritical trust in its outputs.
Learning outcomes
- Multiplatform newswriting and short-form video storytelling.
- Editorial decision-making in AI-assisted workflows.
- Technical execution of short-form video in AI-assisted workflows.
- Reflective and ethical AI use (AI literacy).
Assessment
Assessment considered prompt-based scripting, AI-assisted production outputs, and reflective AI Diaries documenting iterations, editorial decisions, accuracy checks, and ethical considerations. The evaluation prioritised the learning process, students' reflective capacity about AI use, and the development of AI literacy, rather than focusing primarily on the quality of the final output.
Evaluation
Through observation of engagement and learning during activities, analysis of student outputs (descriptive and reflective AI Diaries, scripts, videos), questionnaires on perceptions, class discussions, and quality of the final output.
- KPIs tracked: No
- Formal institutional evaluation: Yes
Risk management
Key risks were tied to journalism-specific concerns, including over-reliance on AI and uncritical acceptance of outputs. Mitigation strategies included explicit 'human-in-the-loop' framing, mandatory AI Diaries, verification requirements, ethical guidelines, transparency discussions, and process-focused assessment. Limited access to paid tools was managed by adapting tasks to free versions and emphasising learning objectives over production quality.
Challenges
- Students showed strong functional dependence on ChatGPT — using it as a trigger or substitute for their own thinking — while paradoxically lacking basic understanding of how the tool works, including the concept of prompting.
- Addressed through foundational AI literacy sessions, structured prompting guides, and reflective AI Diaries. Continuous reminders reinforced verification and human oversight.
- A relational dimension to over-reliance: some students interacted with ChatGPT as if it were a person or 'friend,' leading to uncritical trust in its outputs without fully exploiting its functional potential.
- Managed through ethical framing, structured reflection, and class-wide discussion of AI as a tool, not a collaborator.
- Limited access to paid tools, especially for video production.
- Tasks were adapted to free versions, and assessment was reframed to emphasise process over final output quality.
Scalability
The approach can be transferred to other journalism and media courses. Scaling requires institutional support for tool access (ideally paid versions for video), teacher development in AI literacy, shared prompt and ethics guidelines, and time for reflective components. Curriculum alignment is needed to embed human-in-the-loop workflows and process-based assessment.