Teaching & Learning
- Submitting AI-produced content as one's own without acknowledgement is treated as academic misconduct
- For assessed coursework, the university's default position is that students must not use generative AI unless that use is explicitly authorised
- Schools and module coordinators can permit or encourage AI use for particular assessments, and those permissions must be stated in assessment instructions or handbooks
o Students should not use Generative AI for the preparation and presentation of assessed work unless authorised explicitly by the module co-ordinator. If AI use is permitted, the
The University’s default position regarding students who submit content produced by Generative AI as their own work is as follows:
• If a student submits content produced by Generative AI as their own work without acknowledgement, this will be considered academic misconduct.
5. Any instances in which Generative AI may be used for assessed work should be clearly specified in the assessment instructions and/or School handbook.
6. Each School is responsible for producing student guidelines about the use of Generative AI in assessment within their specific discipline(s), e.g. in School handbooks and/or module handbooks.
- In examinations or class tests, electronic devices, software, or materials are prohibited unless specifically permitted
- For assessments, AI use is governed by school, module, or assessment-specific instructions, and students must follow lecturer or tutor guidance on whether generative AI is permitted for a particular piece of assessment
Schools will communicate to students when the use of AI is acceptable. This will depend on the specific intended learning outcomes associated with the modules/assessments. To avoid this type of academic misconduct, students should always follow the guidance of their lecturer or tutor, who will explain whether the use of generative AI is permitted for a particular piece of assessment.
Academic misconduct in examinations or class tests includes the following prohibited activities:
• Taking electronic devices, software or materials into an examination venue (other than those specifically permitted), irrespective of whether or not any use of the item(s) was made.
- Even where it is allowed, the university says students are still expected to develop study skills they can use independently of AI
- Students may use generative AI for notes, study aids, and other learning materials, but this permission is not uniform because specific modules may disallow it
• Students may choose to use Generative AI to generate notes, study aids, or other materials that they consider helpful in their learning. This type of usage is not prohibited.
o While students may choose to use Generative AI to generate notes, study aids, or other materials that they consider helpful in their learning, this type of usage may not be allowed in specific modules, so students should always check the relevant School or module handbook. Even when this type of usage is allowed, students should be mindful that they are still expected to develop relevant study skills they can employ independently of AI tools.
- The university does not define a separate coding-specific AI policy
- Programming-related AI use falls under the general assessment rule that generative AI is not allowed for assessed work unless explicitly permitted by the module coordinator or school guidance
o Students should not use Generative AI for the preparation and presentation of assessed work unless authorised explicitly by the module co-ordinator. If AI use is permitted, the
Whether Schools permit or encourage the use of Generative AI in your assessments will depend on the specific intended learning outcomes associated with the modules/assessments.
Research
- For postgraduate research students, any use of generative AI in thesis writing beyond the general limited allowance must be discussed and agreed with the supervisor before use, and the use must also be highlighted in annual review
- Any AI use must be acknowledged, cited, and documented with prompts and responses in an appendix, and the university expects such use to be limited to specific sentences or paragraphs rather than broader contribution to the thesis
7. PGR students considering the use of Generative AI that goes beyond what is permitted in principle 2 above should discuss and agree appropriate scope, boundaries, and application with their supervisor at the start of the project or before commencing usage. Any Generative AI usage should also be highlighted during the annual review process and discussed with the review committee.
8. Any use of Generative AI should be acknowledged.
a) AI use should be cited, and any prompts used and responses received reported in an appendix. It is expected that Generative AI, even when used, will be applied only to specific sentences or paragraphs, and not used more broadly to contribute to the thesis.
Extensive usage will still be considered poor academic practice or academic misconduct and investigated accordingly.
b) Use of AI specifically for language correction should be acknowledged in the same way as any other assistance with writing is acknowledged as per the Language Correction policy.
- The university does not define an AI-specific policy for using AI in research data collection or analysis
- However, its academic and research misconduct policies prohibit fabrication, falsification, deliberate omission of data, and false citation in research-related work
Falsification is the fabrication or alteration of data – for example, by changing data in order to confirm a hypothesis not supported by the actual data, or the invention or fabrication of the results of an experiment, which are then reported as genuine measurements. Included in falsification is the deliberate omission of data where, for example, experimental results or known facts are omitted in order to support an otherwise unsupportable hypothesis.
False citation is the citing of a source for information when the source does not contain that information or when the information cited was not gleaned from that source.
ii. Fabrication and/or falsification in proposing (including applications for funding), carrying out or reporting the results of research (including reporting to research funders). This includes explicit and implicit misrepresentation of credentials, qualifications and/or experience.
- It states that research misconduct includes plagiarism and fabrication or falsification in proposing, carrying out, or reporting research, and that the university operates in line with the Concordat to Support Research Integrity
- The university does not set out an AI-specific research ethics framework in these sources, but it does apply research integrity standards across the research lifecycle, including funding applications, planning, undertaking, documenting, communicating, and reviewing research
14.This definition is based on existing norms and best practice. Committing one of the following acts, via commission or omission, as a result of intention and/or recklessness and/or gross negligence, in any aspect of the research endeavour (e.g. obtaining funding, planning research, undertaking research, documenting and communicating the findings, reviewing the work of others)1
i. Misappropriation of the materials, resources or work of others, including plagiarism, misquoting, taking undue credit or the unethical use of privileged material (for example, material seen in reviewing, examining or refereeing).
ii. Fabrication and/or falsification in proposing (including applications for funding), carrying out or reporting the results of research (including reporting to research funders). This includes explicit and implicit misrepresentation of credentials, qualifications and/or experience.
The University of St Andrews fully supports the principles laid out in the UK Concordat to Support Research Integrity (Universities UK, updated 2019): this report summarises our approach to embedding research integrity in our activities, including ensuring compliance with the provisions of the Concordat, and relevant activities undertaken.
Academic Integrity
- When AI use is authorised, students are required to acknowledge it
- For postgraduate research students, AI use must be cited and prompts and responses reported in an appendix
- The university instructs students to include a declaration statement with their references naming the tool, describing how it was used, and giving the access date; it also requires prompts to be retained and made available on request
o If AI use is authorised, all prompts used should be retained for reference and should be available to the module co-ordinator on request.
7. How do I acknowledge use of generative AI?
o You should include a declaration statement along with your references stating the AI tool used, describing how you used it, and indicating the date you accessed it.
o Further requirements may be stipulated by a School, academic programme, or individual
8. Any use of Generative AI should be acknowledged.
a) AI use should be cited, and any prompts used and responses received reported in an appendix.
- Undisclosed submission of AI-produced content as one's own is treated as academic misconduct
- The staff guidance says Turnitin's AI-detection function has not been enabled at St Andrews because of concerns about accurate detection, and the university keeps plagiarism detection reports where applicable as supporting evidence in cases
• If a student submits content produced by Generative AI as their own work without acknowledgement, this will be considered academic misconduct.
A number of plagiarism detection tools to identify the use of generative AI have been reported, including Turnitin. Whilst Turnitin has developed AI-detection, this has not currently been switched on at the University of St Andrews due to concerns over accurate detection.
c) copy of the plagiarism detection software report (if applicable) and any supporting evidence provided by the AMO (this should also be sent to the student(s)).
Institutional & Administrative
- The sources provided do not define a broader policy for staff use of AI in grading, recommendation letters, lesson planning, or administrative communications
- Staff are responsible for communicating to students when AI use is acceptable in assessments, based on intended learning outcomes, and for explaining whether generative AI is permitted for particular assessments
Schools will communicate to students when the use of AI is acceptable. This will depend on the specific intended learning outcomes associated with the modules/assessments. To avoid this type of academic misconduct, students should always follow the guidance of their lecturer or tutor, who will explain whether the use of generative AI is permitted for a particular piece of assessment.
- The university's information security policy applies to staff and students and prohibits entering university information classified as internal, confidential, or strictly confidential into publicly available generative AI tools
- It also says individuals should not enter information that could cause damage if compromised or lost, warns about copyright and personal data risks, and requires reporting to the university if personal data may have been passed to a publicly available generative AI tool
Use of University personal data and other information with Generative Artificial Intelligence (GenAI)
1. Information that is classified as internal, confidential, or strictly confidential by the University Information Classification Policy must not be disclosed/input into publicly available Generative AI tools.
2. Individuals should not put information into publicly available Generative AI tools that, if compromised or lost, could have damaging consequences for other individuals, groups of individuals, or the University more generally.
a. This could also include copyright protected materials, where the volume of material used in Generative AI tools exceeds the provisions for fair dealing, available to educational institutions and students, thus breaching copyright, unless a licence is secured to use protected materials for that purpose.
If personal data is thought to have been passed to a publicly available Generative AI tool, then email dataprot@st-andrews.ac.uk, with the details, as such may be a personal data breach.
- The sources do not provide a university-wide AI strategy or roadmap
- They do show governance structures relevant to research integrity, including a Sub-Committee on Research Integrity, and they state that the university supports the Concordat to Support Research Integrity and is focused on continuous improvement in embedding research integrity
The University of St Andrews fully supports the principles laid out in the UK Concordat to Support Research Integrity (Universities UK, updated 2019): this report summarises our approach to embedding research integrity in our activities, including ensuring compliance with the provisions of the Concordat, and relevant activities undertaken.
As articulated in annual statements, the University is already compliant with the core requirements of the Concordat. However, the University’s approach is to focus on continuous improvement activities aimed at culture-building, taking an academic-led approach to ensure that we strategically focus on activities with a high likelihood of impact.
Owner: Sub-Committee on Research Integrity
Delivery vehicle: Project undertaken by Sub Committee on Research Integrity
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 St Andrews's AI Policies
University of St Andrews has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
When AI use is authorised, students are required to acknowledge it. The university instructs students to include a declaration statement with their references naming the tool, describing how it was used, and giving the access date; it also requires prompts to be retained and made available on request. For postgraduate research students, AI use must be cited and prompts and responses reported in an appendix.
Undisclosed submission of AI-produced content as one's own is treated as academic misconduct. The staff guidance says Turnitin's AI-detection function has not been enabled at St Andrews because of concerns about accurate detection, and the university keeps plagiarism detection reports where applicable as supporting evidence in cases.
The university's information security policy applies to staff and students and prohibits entering university information classified as internal, confidential, or strictly confidential into publicly available generative AI tools. It also says individuals should not enter information that could cause damage if compromised or lost, warns about copyright and personal data risks, and requires reporting to the university if personal data may have been passed to a publicly available generative AI tool.