Core Principles for AI‑Assisted Marking and Feedback at Coventry University Group
Guidance
Purpose of this Guidance
This guidance outlines the appropriate, ethical, and transparent use of Artificial Intelligence (AI) to support marking and feedback at Coventry University Group. It reflects the Group’s position that AI may enhance academic practice but must not replace professional academic judgement or responsibility.
AI‑assisted marking and feedback:
- must remain human‑led
- should support student learning and understanding
- must align with Coventry University Group’s regulations, academic integrity principles, and data protection requirements.
AI must never independently determine marks, grades, or progression decisions. These responsibilities remain strictly with academic staff.
How AI‑Assisted Feedback Can Support Academic Practice
AI tools may be used to support staff effectiveness and feedback quality, provided that academic staff retain oversight and responsibility. Appropriate uses include:
- Enhancing and refining feedback language
Using AI to clarify meaning, improve tone, reduce jargon, or increase accessibility in feedback originally authored by academic staff. - Aligning feedback language and phrasing to assessment criteria thresholds
Mapping marker feedback notes to rubrics to support consistent feedback language aligned to criteria thresholds. - Improving accessibility of feedback formats
Converting written feedback to audio, or summarising spoken feedback (e.g. live assessments). - Synthesising cohort‑level feedback
Summarising patterns across individual feedback to produce coherent whole‑cohort or feed‑forward comments for teaching enhancement. - Supporting rubric development
Using AI to generate draft rubrics aligned to module and course learning outcomes, subject to academic review and approval.
In all cases, AI outputs must be reviewed, edited, and contextualised by the academic marker before release to students.
Staff and Student Trust in AI‑Assisted Marking and Feedback
Research highlights that while AI‑supported marking and feedback can be valued for its clarity, accessibility, and immediacy, students report higher trust in human academic feedback due to its contextual knowledge, personalisation, and disciplinary expertise.
Key factors influencing trust include:
- perceived authenticity of summative and formative feedback
- consistency of AI use for feedback across modules and courses
- transparency about when and how AI is used for marking and feedback
- staff confidence and competence in using AI appropriately
Inconsistent practices risk creating confusion or perceptions of unfairness. Coventry University Group staff should therefore adopt clear, shared, and transparent approaches to AI‑assisted marking and feedback.
Operational Guidelines for Academic Staff
Academic staff are expected to:
- Use AI safely, securely, ethically, and proportionately
- Use Coventry University Group‑approved tools that meet GDPR and data‑protection standards
- Not upload identifiable student work or personal data into AI tools without explicit permission
- Follow University Group, and local entity guidance on AI use
- Reflect on how AI‑assisted feedback may affect student perceptions of fairness, authenticity, and inclusion
- Seek and respond to student feedback on AI‑supported practices
AI must never independently determine marks, grades, or progression decisions. These responsibilities remain strictly with academic staff.
Alignment with Coventry University Group Assessment Principles
This guidance supports Coventry University Group’s commitment to:
- maintaining academic integrity in an AI‑enabled environment
- designing assessments, such as process focused assessments, that foreground student learning and application
- promoting transparent, educative conversations about AI use
- ensuring due process, equity, and quality assurance in marking and feedback
AI‑assisted marking and feedback should be viewed as one tool within reflective, scholarly assessment practice, not as an efficiency shortcut or automated decision‑maker.
References
AIniHE (2025) Riding the tiger of AI Feedback https://aiinhe.org/wp-content/uploads/2025/03/2025-hedx-feedback-survey-insights.pdf
Nazaretsky, N., et al (2024) AI or human? Evaluating student feedback perceptions in higher education. 19th European Conference on Technology Enhanced Learning DOI: 10.1007/978-3-031-72315-5_20
Ofqual (2026) Principles of AI use in marking
This guidance draws upon guidance developed by Kings College London.
