Abstract:

Automated feedback generated by artificial intelligence is increasingly available to university students, yet teachers frequently report two persistent classroom challenges: students can feel overwhelmed by the volume of comments, and feedback is often not translated into concrete revisions. This presentation reports on a classroom innovation designed to convert feedback into purposeful student action within a multimodal opinion/feature article assignment.

The innovation introduces a short triage‑and‑revision plan routine inserted between receiving feedback and revising a draft. Students categorise feedback according to priority, select a small number of high‑impact issues, and translate these into specific revision actions linked to assignment requirements such as reader engagement, evidence integration through hyperlinks, and multimodal design. Revision plans are submitted through a structured form, which is automatically aggregated and summarised using an AI‑supported analysis tool to surface patterns across the cohort. This summary supports teacher awareness of common issues and informs targeted follow‑up support, including brief mini‑lessons, focused peer‑review prompts, or short consultations.

To address concerns about over‑automation and loss of teacher judgement, AI is used solely for pattern recognition and workload reduction rather than for evaluating individual student work or determining instructional responses. Teachers retain full control over interpretation and pedagogical decision‑making, and students are explicitly encouraged to accept, adapt, or reject automated feedback with justification. The presentation shares ready‑to‑use materials and discusses practical considerations for implementing the approach at scale while maintaining student voice, academic integrity, and teacher oversight.

Keywords:

Automated feedback, Teacher‑in‑the‑loop, Responsible AI use, Scalable feedback workflows.