Teaching & Learning
- Use of generative AI in coursework is at instructor and assessment-design discretion rather than uniformly allowed or banned
- The university states that some assessments may permit AI, but using AI to complete work that is submitted as a student's own original work can be academic misconduct if not clearly authorised and acknowledged
Different approaches have different strengths and weaknesses and may not be suitable for all assessments.
Permissive approaches to assessment could, for example, encourage students to use technology as part of the process of producing an assignment, where they use the outputs in a creative, critical or reflective way, acknowledge this process and produce an original submission.
The use of any information generated by AI tools may amount to academic misconduct if submitted as a student's own original work.
Be clear about whether and how students can use GenAI in assessments and other learning activities.
- The university does not impose a single rule for all assessments; approaches may be prohibitive, permissive, or trust-based depending on the assessment
- Staff are expected to state clearly whether and how AI may be used in assessments, and AI-generated material submitted as original work may constitute academic misconduct
Assessments can be designed with:
• a prohibitive approach, i.e. students are not allowed to use technology in a way that directly modifies or generates content;
• a permissive approach, i.e. students are allowed to use technology to support certain aspects of the assessment task with appropriate acknowledgement;
• a trusting approach i.e. students can use technology as they wish but are expected to acknowledge all use of GenAI.
Be clear about whether and how students can use GenAI in assessments and other learning activities.
The use of any information generated by AI tools may amount to academic misconduct if submitted as a student's own original work.
- Students may use generative AI as a learning support tool for studying, idea generation, explanation, and feedback, but they are expected to use it critically and ethically
- The university advises students to check outputs carefully, avoid overreliance, and remember that permitted study support does not automatically translate into permission to use AI in assessed work
You can use AI to support your studying, for example to:
brainstorm ideas
explain concepts in a different way
create practice questions or quizzes
summarise your notes
check grammar or structure in drafts
Like any source, however, GenAI can make mistakes or provide misleading information, so it is important to use these tools critically and thoughtfully. You should not rely on GenAI to do your thinking for you.
Do not assume that because a GenAI tool is available, its use is automatically allowed in your course or assessments. Always check the guidance from your department/module and your tutor.
Research
- Researchers must not list AI tools as authors because authorship requires responsibility that only humans can take
- If generative AI is used in preparing research outputs, the human authors remain fully responsible and should acknowledge that use where appropriate under publisher or disciplinary requirements
Researchers should note that AI tools cannot be listed as authors on research outputs, as authorship requires accountability and responsibility which only humans can assume.
Any use of AI in the preparation of research outputs should be appropriately acknowledged where required, and authors remain responsible for the accuracy, integrity and originality of their work.
- The university requires researchers to use AI in research with care for accuracy, integrity, confidentiality, and legal compliance
- Researchers remain responsible for outputs and must ensure that use of AI in handling or analysing research material does not breach data protection, confidentiality, intellectual property, or ethical obligations
The use of artificial intelligence (AI) and automated tools in research must be undertaken responsibly and with due regard to accuracy, confidentiality, intellectual property, data protection and ethical approval requirements.
Researchers remain responsible for all aspects of their research, including where AI or automated tools have been used.
- Durham explicitly treats responsible AI use as part of research integrity
- AI use in research must comply with ethical approval, confidentiality, data protection, and intellectual property requirements, and misuse or misleading presentation involving AI can fall within research misconduct processes
The use of artificial intelligence (AI) and automated tools in research must be undertaken responsibly and with due regard to accuracy, confidentiality, intellectual property, data protection and ethical approval requirements.
Researchers remain responsible for all aspects of their research, including where AI or automated tools have been used.
Research misconduct may include, but is not limited to: fabrication; falsification; plagiarism; deception in proposing, carrying out or reporting the results of research; failure to comply with legal, ethical and professional obligations; and misrepresentation of data, authorship or interests.
Academic Integrity
- Disclosure requirements depend on context, but the university repeatedly requires acknowledgement when AI use is permitted or expected
- In assessments under permissive or trust-based approaches, students are expected to acknowledge AI use, and in research outputs AI use should be acknowledged where required
Permissive approaches to assessment could, for example, encourage students to use technology as part of the process of producing an assignment, where they use the outputs in a creative, critical or reflective way, acknowledge this process and produce an original submission.
• a trusting approach i.e. students can use technology as they wish but are expected to acknowledge all use of GenAI.
Any use of AI in the preparation of research outputs should be appropriately acknowledged where required, and authors remain responsible for the accuracy, integrity and originality of their work.
- The university frames undisclosed or unauthorised AI-generated material submitted as original student work as potential academic misconduct
- It also cautions staff against overreliance on AI-detection tools, indicating that such tools are unreliable and should not be used as sole evidence in misconduct decisions
The use of any information generated by AI tools may amount to academic misconduct if submitted as a student's own original work.
AI detection tools are not sufficiently reliable to be used as sole evidence of misconduct.
Markers should not assume AI use based solely on writing style or on outputs from detection software.
Institutional & Administrative
- Staff should not rely on AI-detection outputs or automated judgments alone when assessing possible misconduct
- Faculty are encouraged to engage with generative AI in teaching and assessment design, but they must set clear rules for students and exercise human judgment in marking
We encourage colleagues to explore how generative AI can support teaching, learning and assessment, while being mindful of the challenges and risks.
Be clear about whether and how students can use GenAI in assessments and other learning activities.
Markers should not assume AI use based solely on writing style or on outputs from detection software.
AI detection tools are not sufficiently reliable to be used as sole evidence of misconduct.
- Durham requires care with privacy, confidentiality, and data protection when using AI, including in research
- The provided sources do not identify a university-approved list of AI platforms, but they do state that AI use must respect data protection and confidentiality obligations
The use of artificial intelligence (AI) and automated tools in research must be undertaken responsibly and with due regard to accuracy, confidentiality, intellectual property, data protection and ethical approval requirements.
Like any source, however, GenAI can make mistakes or provide misleading information, so it is important to use these tools critically and thoughtfully.
- Its published approach emphasises exploration of generative AI in teaching, learning, and assessment alongside attention to integrity and assessment redesign
- The university presents an institution-level educational AI guidance approach focused on supporting use while managing risks, rather than a single blanket rule
This page provides Durham University staff and students with guidance and resources on the use of generative artificial intelligence (GenAI) in teaching, learning and assessment.
We encourage colleagues to explore how generative AI can support teaching, learning and assessment, while being mindful of the challenges and risks.
The emergence of generative AI means that any assessment can potentially be completed by AI, but this is also true of other forms of support students have always had available. The challenge is to think carefully about what we want students to learn and demonstrate, and then to design assessments accordingly.
DocuMark: Responsible AI Use for Academic Integrity
Knowing your institution's AI policy is step one. DocuMark helps enforce it fairly by empowering universities to manage AI-generated content, prevent cheating, and support student writing through responsible AI use.
Common Questions About Durham University's AI Policies
Durham University has defined AI policies in 11 of 12 categories, with an overall coverage score of 92%.
Disclosure requirements depend on context, but the university repeatedly requires acknowledgement when AI use is permitted or expected. In assessments under permissive or trust-based approaches, students are expected to acknowledge AI use, and in research outputs AI use should be acknowledged where required.
The university frames undisclosed or unauthorised AI-generated material submitted as original student work as potential academic misconduct. It also cautions staff against overreliance on AI-detection tools, indicating that such tools are unreliable and should not be used as sole evidence in misconduct decisions.
Durham requires care with privacy, confidentiality, and data protection when using AI, including in research. The provided sources do not identify a university-approved list of AI platforms, but they do state that AI use must respect data protection and confidentiality obligations.