Teaching & Learning
- Some assignments may also require a declaration describing how AI was used
- Students may use generative AI to support learning, but they must not submit work created or part-created by AI as their own
- Use in assessed work can be restricted by learning outcomes, and the university directs students to check with their tutor because acceptable use may vary by task
You are allowed to use GenAI tools like ChatGPT to develop your learning and support your studies. However, you should never submit work that has been created or part-created by GenAI for assessments that you claim as your own work.
This includes situations where you may not be allowed to use GenAI tools because learning outcomes indicate you need to develop and demonstrate your own individual skills and knowledge. In other cases, you may be asked to complete a declaration statement that describes how you have used generative AI. For some learning activities your tutor will build use of GenAI into teaching and give guidelines on how to use the tools.
Please check with your tutor before using generative AI in your work.
* copying and pasting information created by GenAI straight into your work
* asking GenAI to write or rewrite your work or part of your work for you
* directly asking GenAI to answer an assessment question
- The university also treats unauthorized technology use during examinations as cheating
- Whether AI is allowed for a particular assessment depends on the authorized conditions for that assessment
- AI use in assessments is prohibited when students submit AI-created or AI-altered work as their own or use AI in an unauthorized way
You are allowed to use GenAI tools like ChatGPT to develop your learning and support your studies. However, you should never submit work that has been created or part-created by GenAI for assessments that you claim as your own work.
* directly asking GenAI to answer an assessment question
* 1.3.2 Cheating: any action before, during, or after an examination or assessment by which you seek to gain unfair advantage or assist another student to do so. This includes the possession of unauthorised material or technology during an examination and attempting to access unseen assessment materials in advance of an examination.
* 1.3.4 External authorship/assistance: where a student obtains, or attempts to obtain, unauthorised input from another person or service with their work. This may include asking for unauthorised assistance with assessments, or engaging with essay mills, or using artificial intelligence tools in an unauthorised way to generate or alter the content or wording of academic work.
- Students are still advised to check with their tutor before using AI in their work
- The university permits students to use generative AI for non-graded study support and personal learning
- Examples explicitly allowed include idea generation, spelling and grammar feedback, summarising complex texts, and general study support
GenAI tools can be enormously useful in supporting your studies by generating ideas, or by checking or editing work.
You are allowed to use GenAI tools like ChatGPT to develop your learning and support your studies.
Acceptable uses include:
* using GenAI as an assistive technology by arrangement with our Disability and Inclusion team
* using GenAI as a study buddy, for example, giving feedback on some of your writing for spelling or grammar errors, or for getting started with ideas
* citing or clearly referencing your use of GenAI
* summarising complex ideas or academic texts
Please check with your tutor before using generative AI in your work.
- Student use of AI in academic work remains governed by the general rules against unauthorized AI assistance and by tutor or assessment-specific instructions
- The university acknowledges that generative AI can respond to code, but it does not set a distinct university-wide policy for AI code generation in programming assignments
Generative artificial intelligence tools or services can be used to generate text, video, audio and image content in response to questions or instructions, often referred to as prompts. They can give responses to data that is input into them too, such as code or text.
Please check with your tutor before using generative AI in your work.
* 1.3.4 External authorship/assistance: where a student obtains, or attempts to obtain, unauthorised input from another person or service with their work. This may include asking for unauthorised assistance with assessments, or engaging with essay mills, or using artificial intelligence tools in an unauthorised way to generate or alter the content or wording of academic work.
Research
- No more specific university-wide rule on drafting manuscripts with AI is defined in the provided sources
- The policy warns that these tools can introduce bias and inaccuracies and may compromise confidential or sensitive data
- Researchers are told to consider carefully whether generative AI is appropriate for research and to check journal-specific rules before using it in publication-related work
Be aware if the use of Generative Artificial Intelligence (Gen AI) tools are appropriate for your research. It is important to understand that these tools recognise patterns but do not understand the content. Whilst it can lead to biases and inaccuracies in the content they generate, it may breach security of confidential and sensitive data. You will need to check with each journal their rules for using GenAI.
All researchers are expected to publish and disseminate the results of their research in an open, honest, transparent, and accurate manner, and via all appropriate media such as journal papers, books, reviews, software, data repository or conference proceedings.
- No detailed university-wide rule specifically governing AI for data analysis or synthetic data generation is defined in the provided sources
- It also warns researchers that generative AI can produce biased or inaccurate content and may breach the security of confidential and sensitive data
- The university states that AI may be used as a tool that supports research, including studies involving AI-enabled technologies, but research using AI must follow specific ethics principles
In addition to the above general principles, there are specific ethics principles that need to be followed for studies which involve AI, either as the subject of research (like AI-infused medical devices) or in the form of a tool that supports the research (like ChatGPT) or the development of AI-enabled tools / technology.
Be aware if the use of Generative Artificial Intelligence (Gen AI) tools are appropriate for your research. It is important to understand that these tools recognise patterns but do not understand the content. Whilst it can lead to biases and inaccuracies in the content they generate, it may breach security of confidential and sensitive data. You will need to check with each journal their rules for using GenAI.
- In peer review, researchers must not use AI without explicit permission, especially where confidential manuscripts are involved
- The university has explicit ethics principles for research involving AI, whether AI is the subject of study, a research-support tool, or part of an AI-enabled technology
- Researchers deploying AI in research must consider stakeholder-based principles including fairness, accountability, transparency, explainability, reliability, safety, and human autonomy
### 3.8 Principle of ethical conduct of research when using Artificial Intelligence (AI)
In addition to the above general principles, there are specific ethics principles that need to be followed for studies which involve AI, either as the subject of research (like AI-infused medical devices) or in the form of a tool that supports the research (like ChatGPT) or the development of AI-enabled tools / technology.
* End-users: Those who might use the AI-enabled technology (for example, a financial agent or a doctor reliant on a decision-support system) or those who would be on the receiving end of an AI-enabled technology for example, an applicant for a loan, a patient using an automated diagnostic tool) would consider non-discrimination and fairness, accountability, and transparency, explainability and interpretability, reliability and safety and human autonomy.
* Researcher: the person deploying an AI-enabled technology as part of their research protocol - all of those stated above.
Researchers must not use AI to support peer review without explicit permission. Uploading confidential manuscripts into generative tools may breach trust or policy.
Academic Integrity
- The university requires transparency about AI use in student work when requested and treats proper acknowledgement as an acceptable practice
- Students may be required to complete a declaration statement describing their AI use, and the guidance explicitly says AI use should be cited or clearly referenced
In other cases, you may be asked to complete a declaration statement that describes how you have used generative AI.
Acceptable uses include:
* citing or clearly referencing your use of GenAI
* 1.3.1 Plagiarism: the use of ideas, intellectual property, or material derived from another source (including another student’s work, or content generated by any other person or computer software) without appropriate acknowledgement and/or indication, thereby presenting it as your own.
- The provided sources define misconduct and recordkeeping for academic conduct breaches, but they do not define a policy position on AI detection tools such as Turnitin or GPTZero
- Unauthorized AI use is enforceable under the university's academic responsibility and conduct regulations as plagiarism, cheating, or external authorship/assistance, depending on how it is used
* 1.3.1 Plagiarism: the use of ideas, intellectual property, or material derived from another source (including another student’s work, or content generated by any other person or computer software) without appropriate acknowledgement and/or indication, thereby presenting it as your own.
* 1.3.2 Cheating: any action before, during, or after an examination or assessment by which you seek to gain unfair advantage or assist another student to do so. This includes the possession of unauthorised material or technology during an examination and attempting to access unseen assessment materials in advance of an examination.
* 1.3.4 External authorship/assistance: where a student obtains, or attempts to obtain, unauthorised input from another person or service with their work. This may include asking for unauthorised assistance with assessments, or engaging with essay mills, or using artificial intelligence tools in an unauthorised way to generate or alter the content or wording of academic work.
7.2 Records of academic conduct breaches, or where students have been signposted to Additional Learning will be maintained in line with the University's Retention of Assessment Material and Student Records Policy.
Institutional & Administrative
- In teaching, tutors may build AI use into learning activities and provide guidelines
- No explicit policy in the provided sources addresses staff use of AI for grading, recommendation letters, or administrative communications
- The university provides staff guidance on generative AI in education and states that a cross-institutional group has been producing guidance for staff and students
Guidance for staff (links open in SharePoint):
At the University of Southampton, we are actively exploring how AI - and particularly generative AI (GenAI) is transforming education. A cross-institutional group has been working since July 2023 to provide guidance for staff and students, our senior Executive team and to provide a dynamic discussion space for this fast-moving area.
For some learning activities your tutor will build use of GenAI into teaching and give guidelines on how to use the tools.
- Research policies further require confidential information to be managed carefully and not disclosed without consent
- The university identifies Microsoft Copilot within Office365 as an available generative AI tool for students and staff
- It also prohibits students from sharing module materials in generative AI tools and warns researchers that generative AI may breach the security of confidential and sensitive data
Student and members of staff currently have access to the generative AI tool CoPilot within Microsoft Office365.
* sharing the materials and resources from your module tutors in a GenAI tool - you do not have permission to share these materials openly because they belong to the University
Be aware if the use of Generative Artificial Intelligence (Gen AI) tools are appropriate for your research. It is important to understand that these tools recognise patterns but do not understand the content. Whilst it can lead to biases and inaccuracies in the content they generate, it may breach security of confidential and sensitive data.
As a researcher you must:
* manage confidential and/or restricted information with care.
* not use or disclose confidential information to others without the consent of the party who owns it.
- The university also says that one size does not fit all across disciplines
- The university has an institution-level approach to AI in education centered on guidance rather than a single fixed rule for all contexts
- It states that a cross-institutional group has been working since July 2023 to guide staff, students, and senior leadership, and it publishes guiding principles that emphasize collaboration, co-creation, and coordinated development
At the University of Southampton, we are actively exploring how AI - and particularly generative AI (GenAI) is transforming education. A cross-institutional group has been working since July 2023 to provide guidance for staff and students, our senior Executive team and to provide a dynamic discussion space for this fast-moving area.
Our guiding principles in considering the use of AI in education are:
* collaboration - and acknowledgement that ‘one size does not fit all’ with opportunity and challenge varying between disciplines
* co-creation - working with our students on advice, guidance and other activities
* moving forward together - work created by multiple teams across UoS, with expertise led, shared and coordinated through the working group
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 University of Southampton's AI Policies
University of Southampton has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
The university requires transparency about AI use in student work when requested and treats proper acknowledgement as an acceptable practice. Students may be required to complete a declaration statement describing their AI use, and the guidance explicitly says AI use should be cited or clearly referenced.
Unauthorized AI use is enforceable under the university's academic responsibility and conduct regulations as plagiarism, cheating, or external authorship/assistance, depending on how it is used. The provided sources define misconduct and recordkeeping for academic conduct breaches, but they do not define a policy position on AI detection tools such as Turnitin or GPTZero.
The university identifies Microsoft Copilot within Office365 as an available generative AI tool for students and staff. It also prohibits students from sharing module materials in generative AI tools and warns researchers that generative AI may breach the security of confidential and sensitive data. Research policies further require confidential information to be managed carefully and not disclosed without consent.