Marking Guidance (AI)

There is clear evidence and growing guidance emphasizing the importance of balanced, objective, and supportive feedback when AI use is suspected in student work. Current best practice, as outlined by universities and sector bodies, strongly discourages accusatory or punitive language in feedback and instead encourages transparency, clarity, and a focus on learning and academic integrity.

Key Evidence and Guidance

• Avoid Accusatory Tone:
  Universities such as Newcastle and Plymouth explicitly advise against relying on flawed AI detection tools and caution staff not to make accusations based solely on suspicion or detection scores. Instead, any concerns should be handled through established academic misconduct procedures, which emphasize fairness and due process45710.

• Be Transparent and Constructive:
  Guidance from institutions and sector bodies recommends that feedback should:

  • Clearly explain what aspects of the work raised concerns, focusing on observable features (e.g., lack of personal insight, generic content) rather than making direct claims about AI use45.

  • Reference institutional policies and expectations around AI use, so students understand the standards and rationale510.

  • Encourage students to reflect on their learning process and, where appropriate, provide opportunities to clarify or discuss their approach to the assignment510.

• Support Student Learning:
  The focus should be on upholding academic integrity while supporting student development. This includes:

  • Reinforcing the value of the assessment process and the importance of learning outcomes5.

  • Helping students understand the boundaries of acceptable AI use and how to properly acknowledge any assistance1510.

  • Providing resources or guidance for responsible and ethical engagement with AI tools137.

• Objective, Evidence-Based Approach:
  Feedback should be based on evidence within the work itself, not assumptions about AI use. If further investigation is needed, this should follow formal procedures rather than being addressed through feedback alone457.

Example Guidance Language

"We noticed some aspects of your submission that are highly generic and lack specific reference to class discussions or your personal learning journey. Please remember that our policy requires any use of AI tools to be clearly acknowledged, and that your work should demonstrate your own understanding and insights. If you have questions about how to use AI responsibly, please refer to our guidance or arrange a meeting to discuss this further."

Citations:

1. https://altc.alt.ac.uk/blog/2023/10/student-guidance-for-the-responsible-use-of-ai/

2. https://www.gov.uk/government/publications/generative-artificial-intelligence-in-education/generative-artificial-intelligence-ai-in-education

3. https://www.kcl.ac.uk/about/strategy/learning-and-teaching/ai-guidance

4. https://ec.plymouth.ac.uk/ai-considerations-for-staff/

5. https://www.ncl.ac.uk/learning-and-teaching/effective-practice/ai/ai-in-assessment/

6. https://news.stanford.edu/stories/2023/05/ai-feedback-tool-improves-teaching-practices

7. https://www.gcu.ac.uk/__data/assets/pdf_file/0018/173007/Appropriate-use-of-AI_November-2023.pdf

8. https://feedbackfruits.com/blog/ai-and-authentic-assessment

9. https://www.hepi.ac.uk/2024/10/22/how-ai-generated-feedback-could-assist-student-learning/

10. https://mitsloanedtech.mit.edu/ai/teach/ai-detectors-dont-work/

Answer from Perplexity: pplx.ai/share

There is clear evidence and growing guidance emphasizing the importance of balanced, objective, and supportive feedback when AI use is suspected in student work. Current best practice, as outlined by universities and sector bodies, strongly discourages accusatory or punitive language in feedback and instead encourages transparency, clarity, and a focus on learning and academic integrity.

Key Evidence and Guidance

• Avoid Accusatory Tone:
  Universities such as Newcastle and Plymouth explicitly advise against relying on flawed AI detection tools and caution staff not to make accusations based solely on suspicion or detection scores. Instead, any concerns should be handled through established academic misconduct procedures, which emphasize fairness and due process45710.
  
  

• Be Transparent and Constructive:
  Guidance from institutions and sector bodies recommends that feedback should:
  
  

  • Clearly explain what aspects of the work raised concerns, focusing on observable features (e.g., lack of personal insight, generic content) rather than making direct claims about AI use45.
    
    

  • Reference institutional policies and expectations around AI use, so students understand the standards and rationale510.
    
    

  • Encourage students to reflect on their learning process and, where appropriate, provide opportunities to clarify or discuss their approach to the assignment510.
    
    

• Support Student Learning:
  The focus should be on upholding academic integrity while supporting student development. This includes:
  
  

  • Reinforcing the value of the assessment process and the importance of learning outcomes5.
    
    

  • Helping students understand the boundaries of acceptable AI use and how to properly acknowledge any assistance1510.
    
    

  • Providing resources or guidance for responsible and ethical engagement with AI tools137.
    
    

• Objective, Evidence-Based Approach:
  Feedback should be based on evidence within the work itself, not assumptions about AI use. If further investigation is needed, this should follow formal procedures rather than being addressed through feedback alone457.
  
  

Example Guidance Language

"We noticed some aspects of your submission that are highly generic and lack specific reference to class discussions or your personal learning journey. Please remember that our policy requires any use of AI tools to be clearly acknowledged, and that your work should demonstrate your own understanding and insights. If you have questions about how to use AI responsibly, please refer to our guidance or arrange a meeting to discuss this further."

Summary Table

Principle

Guidance for Feedback

Tone

Balanced, objective, non-accusatory

Evidence

Focus on observable features, not unproven suspicion

Clarity

Reference policies and expectations

Support

Encourage learning, offer resources, invite discussion

Process

Use formal academic procedures for suspected misconduct

Conclusion

Sector guidance is clear: feedback around suspected AI use should be fair, balanced, and focused on supporting student learning and integrity, not on accusation or punishment. This approach maintains trust, encourages responsible AI use, and upholds academic standards4510.

Add to follow-up

Check sources

Citations:

1. https://altc.alt.ac.uk/blog/2023/10/student-guidance-for-the-responsible-use-of-ai/

2. https://www.gov.uk/government/publications/generative-artificial-intelligence-in-education/generative-artificial-intelligence-ai-in-education

3. https://www.kcl.ac.uk/about/strategy/learning-and-teaching/ai-guidance

4. https://ec.plymouth.ac.uk/ai-considerations-for-staff/

5. https://www.ncl.ac.uk/learning-and-teaching/effective-practice/ai/ai-in-assessment/

6. https://news.stanford.edu/stories/2023/05/ai-feedback-tool-improves-teaching-practices

7. https://www.gcu.ac.uk/__data/assets/pdf_file/0018/173007/Appropriate-use-of-AI_November-2023.pdf

8. https://feedbackfruits.com/blog/ai-and-authentic-assessment

9. https://www.hepi.ac.uk/2024/10/22/how-ai-generated-feedback-could-assist-student-learning/

10. https://mitsloanedtech.mit.edu/ai/teach/ai-detectors-dont-work/

Answer from Perplexity: pplx.ai/share