University of Portsmouth AI Policy

You must not AI if you have been told not to do so in the assessment brief. You must not use AI to write or complete your assignments for you. Claiming AI-generated work to be your own is academic misconduct. The principle is simple: you must be able to show mastery of what you submit. Your tutors may ask you to discuss your work in person to assess your understanding of what you submitted.

 Our summative assessments should provide credible evidence that each student has achieved learning outcomes, authored their own work, and can demonstrate mastery of the work they submitted

This AI acknowledgement should be part of every asynchronous assessment submission (anything that students produce in their own time). This does not require a change to the Module Descriptor (it is not an assessment artefact), but this requirement must be clearly stated in all coursework briefs.

By the end of the 2027-28 academic year:

1. All colleagues are proficient in a range of AI tools and their pedagogical uses;

2. All assessments are of types that either embrace AI as a legitimate and ethical tool that contributes to learning, or exclude the use of AI, where appropriate, in line with Table 1;

 Track B (in-person only): permitted where aligned with the University’s position on exams.

* Cheating in examinations (or other formal assessments), including possession of unauthorised material or technology during an examination, and attempting to access unseen assessment materials in advance of an examination;

AI can be a great study partner if used wisely. You might use it to:

* Generate ideas or brainstorm topics

* Clarify difficult concepts

* Check grammar or structure

* Explore examples or case studies

* Get feedback on drafts, with your own judgement applied. You may use this feedback to inform and enhance your work. Critically, the work must remain yours, and you should be able to demonstrate full understanding and ownership of what you submit.

Remember: AI should help you learn, not learn for you. Use it responsibly, stay curious, and keep your learning authentic.

1.1.12. Artificial Intelligence: The University permits the use of AI as a tool to assist and inform research and generation of ideas, planning and output. The use of AI in submitted work must adhere to principles of academic integrity, with proper citation and clear indication of where AI has been utilised. Failure to do so will be considered for academic misconduct action. Guidance can be found in the University’s use of AI statement document.

* Data analysis – using AI to analyse data sets or interpret results.

 Maintain data privacy: do not upload personally identifiable information or sensitive data (eg: financial or confidential) into generative AI tools.

 Keep data secure: when using generative AI tools for work, ensure that university business information and data are not placed into the public domain. For example, Copilot is a tool that will not allow data to move outside of UoP infrastructure.

This Policy provides a framework for professional practice and decision-making on ethical issues as they arise in the work of the University within research, innovation, learning and teaching. It also relates to how the University handles research misconduct.

Staff, research students and all others conducting research under the auspices of the University (e.g. University owned spinouts, industrial partners using University facilities etc.) are required to report misconduct in research where they have good reason to believe it is occurring. The University will investigate allegations or complaints about misconduct in research.

• Breaching legal, ethical and professional requirements needed for research, for example those needed for human research participants, animals, or human organs or tissue used in research, or for the protection of the environment. An example of this includes proceeding with research without ethics review or not obtaining informed consent.

• Proceeding with research without necessary permissions and approvals in place.

 Students must acknowledge, not reference, AI use in submissions.

 An acknowledgement on AI use is required with each asynchronous assessment submission.

You must acknowledge whenever you use AI to help with an assessment that you complete in your own time. Each summative assessment dropbox in Moodle will allow you to acknowledge if and how you have used AI to help with your assessment.

The dropbox in Moodle allows students to declare whether generative AI was used and how. Declarations are recorded in Moodle and linked to the submission, so staff can report on AI use and selected categories. Students must select the relevant option before submitting. Submissions are not blocked if it is skipped, but no declaration will be recorded and this can be reported.

1.1.12. Artificial Intelligence: The University permits the use of AI as a tool to assist and inform research and generation of ideas, planning and output. The use of AI in submitted work must adhere to principles of academic integrity, with proper citation and clear indication of where AI has been utilised.

 AI detection tools must not be used, as they are currently unreliable and often misleading.

1.1.12. Artificial Intelligence: The University permits the use of AI as a tool to assist and inform research and generation of ideas, planning and output. The use of AI in submitted work must adhere to principles of academic integrity, with proper citation and clear indication of where AI has been utilised. Failure to do so will be considered for academic misconduct action.

* Improper use of generative AI

Your tutors may ask you to discuss your work in person to assess your understanding of what you submitted.

This guidance sets out the University of Portsmouth’s position on generative AI in learning, teaching and assessment. It encourages staff to educate themselves about and embrace AI rather than attempt to police or ban its use.

 AI tools must not currently be used to upload student work.

 AI detection tools must not be used, as they are currently unreliable and often misleading.

 AI tools must not currently be used for the purposes of marking and feedback.

 Staff can use AI for personal productivity, idea generation, etc., but not with identifiable student data.6

 Be transparent: maintain professional integrity by disclosing when content is AI-generated.

 Check facts: even with the best prompt, AI can make mistakes; verify AI-generated content and linked sources, ensuring it reflects reality and avoids ‘hallucination’.

 Evaluate language and tone: ensure responses are free from language or assumption that could marginalise any group and use individual judgement to review appropriateness for work.

 Avoid AI for critical decision-making: AI should not be used for making decisions that significantly impact people, such as recruitment, promotion, or disciplinary action.

 Maintain IP oversight: ensure human oversight and refinement of AI produced content to track and verify source provenance and meet copyright requirements

AI is here to help. Used well, it saves time, cuts down on mundane work, and allows staff to focus on what really matters. The key is getting used to using it and learning how to use it most effectively and ethically. This step-by-step journey starts with experimentation, so staff are encouraged to try it out using our secure, Microsoft AI tool, Copilot.

 Maintain data privacy: do not upload personally identifiable information or sensitive data (eg: financial or confidential) into generative AI tools.

 Keep data secure: when using generative AI tools for work, ensure that university business information and data are not placed into the public domain. For example, Copilot is a tool that will not allow data to move outside of UoP infrastructure.

 Staff can use AI for personal productivity, idea generation, etc., but not with identifiable student data.6

This step-by-step journey starts with experimentation, so staff are encouraged to try it out using our secure, Microsoft AI tool, Copilot.

Statement of intent

By the end of the 2027-28 academic year:

1. All colleagues are proficient in a range of AI tools and their pedagogical uses;

2. All assessments are of types that either embrace AI as a legitimate and ethical tool that contributes to learning, or exclude the use of AI, where appropriate, in line with Table 1;

3. Students are confident in their use of AI;

4. The University has an academic conduct environment that supports student development and understanding of the use of AI tools.

Core principles:

1. Educate and embrace AI use – don’t police or ban it

2. AI should enhance, not replace academic practices or roles

3. Focus on learning and authenticity, not detection or suspicion

4. (Co-)design assessments to integrate AI transparently and ethically

UoP is committed to sustainable practices and will select technology and companies with responsible environmental policies and will establish clear guidelines and training relating to standards and practices in the use of AI, to ensure transparency, fairness, and accountability.