Teaching & Learning
- For assessed student work, AI use is not universally allowed or banned; it depends on the rules for the specific assessment
- Students must not present AI-generated work or ideas as their own, and if AI is allowed in that assessment, they must fully cite its use
6.1 Academic misconduct is defined as an attempt to obtain, obtaining, or assisting another person in obtaining an unfair advantage in an academic assessment. An offence may occur in relation to any form or component of assessment, including but not limited to coursework, examinations, research proposals and reports, presentations, posters and digital media works.
h. Improper use of AI: Students shall not claim work or ideas that have been generated through Artificial Intelligence as their own; where AI usage is permitted under the guidelines of the relevant assessment, its usage must be fully cited in accordance with academic best practice on referencing.
- AI use in examinations and other assessments is governed by whether permission is given in the relevant assessment rules
- Using AI without permission, or presenting AI-generated material as one's own, is treated as academic misconduct; for doctoral researchers, AI use in the viva voce examination is prohibited
6.1 Academic misconduct is defined as an attempt to obtain, obtaining, or assisting another person in obtaining an unfair advantage in an academic assessment. An offence may occur in relation to any form or component of assessment, including but not limited to coursework, examinations, research proposals and reports, presentations, posters and digital media works.
h. Improper use of AI: Students shall not claim work or ideas that have been generated through Artificial Intelligence as their own; where AI usage is permitted under the guidelines of the relevant assessment, its usage must be fully cited in accordance with academic best practice on referencing.
The use of AI in the viva voce examination is prohibited and doctoral researchers should consider very carefully the potential risks of using AI in assessed work.
Research
- The university permits AI in research and knowledge exchange, including to support academic writing, but it must not replace the researcher's independent and original thought, critical analysis, and writing
- Researchers must acknowledge AI use, check AI-generated outputs for errors and inconsistencies, and clearly reference AI use in reports, publications, and doctoral theses; doctoral students are advised to seek supervisor advice before using AI in research and writing
Artificial intelligence (AI), including generative AI, can have many useful applications in academic research, including initial searches for ideas, data analysis and visualisation, research design and development and supporting academic writing.
Roehampton therefore permits the use of artificial intelligence in research and knowledge exchange activity. However, AI must not be used in a way which would undermine academic integrity, hinder researcher development or substitute for the researcher’s independent and original thought, critical analysis and writing.
Transparent: Researchers must clearly acknowledge the use of AI in their research and failure to do so will be regarded as an example of academic and/or research misconduct. To support this, researchers should keep records of their AI use, including prompts employed and responses received. References should make clear the AI tool(s) used, who inputted the data and or prompt and the date of use.
Doctoral researchers should seek the advice of their supervisors on using AI before employing it in their research and writing.
Reference the use of AI in research clearly in any outputs generated from the research, including reports, publications and doctoral theses.
AI tools can be helpful in improving the clarity and structure of academic writing, including doctoral theses.
Researchers must check any AI-generated outputs for errors and inconsistencies and should be transparent regarding their use of AI in their
- Researchers should keep records of AI use in data analysis and transcription, check outputs carefully for errors and bias, and must not falsify or fabricate data through AI manipulation
- The university permits AI for research data analysis and related research activity, but researchers remain individually accountable and must use it ethically, responsibly, and transparently
Artificial intelligence (AI), including generative AI, can have many useful applications in academic research, including initial searches for ideas, data analysis and visualisation, research design and development and supporting academic writing.
Roehampton therefore permits the use of artificial intelligence in research and knowledge exchange activity.
Researchers, including doctoral students at the University, are therefore required to ensure that their use of AI in research remains ethical, responsible and transparent with accountability linking to the individual researcher.
- Ensure that records of AI use in data analysis, transcription etc. are kept, including the specific tool and version, inputs and outputs, and date of use.
- Check AI-generated content carefully, being attentive not only to errors/’hallucinations’ but also to possible bias in AI-produced material.
• Falsify or fabricate data through the manipulation of material, including images via AI
- Roehampton requires AI use in research to meet ethical standards, comply with data protection and third-party rights, and obtain formal ethical approval where necessary
- The university treats non-transparent AI use as academic or research misconduct, expects research students to maintain academic integrity and honesty, and has committee oversight for ethics approval processes, including high-risk applications
Ethical: Researchers are required to ensure that their use of AI is within accepted ethical standards.
Use of AI must adhere to data protection laws and avoid the infringement of rights of third parties. Formal, ethical approval must be sought where necessary for the use of AI. Advice can be sought from the School’s ethics officer if you are unclear.
Transparent: Researchers must clearly acknowledge the use of AI in their research and failure to do so will be regarded as an example of academic and/or research misconduct.
The university also expects that research students will maintain principles of academic integrity and honesty when presenting or disseminating their research, either at internal or external events or through publication.
Oversee, and report on, the Ethics approval process, including scrutiny and approval or otherwise of high-risk Ethics applications
Academic Integrity
- Disclosure of permitted AI use is mandatory for both student assessed work and research outputs
- Students must fully cite AI use when it is permitted in an assessment, while researchers must clearly acknowledge AI use, keep records of prompts and responses, identify the tool and date used, and provide acknowledgment or methodological explanation where appropriate
h. Improper use of AI: Students shall not claim work or ideas that have been generated through Artificial Intelligence as their own; where AI usage is permitted under the guidelines of the relevant assessment, its usage must be fully cited in accordance with academic best practice on referencing.
Transparent: Researchers must clearly acknowledge the use of AI in their research and failure to do so will be regarded as an example of academic and/or research misconduct. To support this, researchers should keep records of their AI use, including prompts employed and responses received. References should make clear the AI tool(s) used, who inputted the data and or prompt and the date of use. Acknowledgements or longer methodological explanations should be provided where
Reference the use of AI in research clearly in any outputs generated from the research, including reports, publications and doctoral theses. Where AI is integral to the research design of a project, this should be clearly explained in the discussion of the project methodology. General acknowledgment should be given where AI has been used to support academic writing and the preparation of outputs.
AI tools just as they may also make use of human proof-readers, though as with human proof-readers that use must be acknowledged.
- Undisclosed or improper AI use is enforceable as academic misconduct
- Research students submitting theses must use Turnitin, include an originality report, and work with supervisors if the similarity percentage is above 20%
- For taught students, misuse of AI carries sanctions up to permanent expulsion depending on level and prior offences; for postgraduate research students, misuse of AI can lead to permanent expulsion or required remediation
Transparent: Researchers must clearly acknowledge the use of AI in their research and failure to do so will be regarded as an example of academic and/or research misconduct.
UG and PG Taught
Plagiarism,
duplication,
falsification,
collusion,
cheating,
misuse of AI
0 2 Mark of 0 for work,
resit required for
work to be capped at
pass
2 3 Permanent expulsion
PG Research
Plagiarism,
duplication,
falsification,
collusion,
cheating,
misuse of AI
0 3 Permanent expulsion
Student required to
remedy elements of work
produced through
misconduct
In addition to completing the RDCom10 (Candidate’s Declaration Form), students will also need to submit their thesis via the Turnitin Dropbox on the Graduate School Moodle site and download the originality report to include with their RDCom10.
If students return a similarity percentage of higher than 20%, they should work with supervisors to address any issues raised by the report before submitting the thesis.
Cases of suspected academic misconduct, whether plagiarism, cheating, collusion (unauthorised cooperation), falsification, contract cheating, inappropriate use of artificial intelligence or otherwise, will be investigated under the terms of the university’s Student Disciplinary Regulations. In the severest cases, students may be expelled from the university on the grounds of academic misconduct.
Institutional & Administrative
- Staff engaged in certain human research procedures must attend GCP training, and staff researchers are covered by the research AI policy
- The sources define staff obligations for research use of AI and for certain research training, but they do not define a general policy for faculty use of AI in grading, feedback, lesson planning, recommendation letters, or administrative communications
This half-day core course is intended for staff and PhD researchers who plan to undertake work involving human research procedures at the University of Roehampton.
Attendance is therefore required for all staff and PhD students engaged in this type of research.
Researchers, including doctoral students at the University, are therefore required to ensure that their use of AI in research remains ethical, responsible and transparent with accountability linking to the individual researcher.
- Roehampton requires university information to be classified and handled according to its classification level, which affects what may be entered into or stored on external tools
- Internal-use information must stay on university-managed systems, restricted and confidential information must only be held on approved secure systems, and confidential information must use university-approved encrypted transfer methods; researchers are also advised to use AI tools that do not retain or train on user inputs, and staff/students must not store protected information in personal cloud services unless securely encrypted first
5.1 All University data should be classified in line with the following four information classification levels:
Public – Information intended for unrestricted public access with no anticipated harm arising from disclosure.
Internal Use – Information for use within the University community where unauthorised disclosure could cause minor inconvenience or reputational impact.
Restricted – Information where unauthorised disclosure may result in significant harm, including legal, financial, operational, or reputational damage.
Confidential – Information requiring enhanced controls due to the substantial harm that may result from unauthorised disclosure, including serious legal or regulatory consequences.
Information must be held on University‑managed systems, not on public platforms
Information should only be held on approved secure systems, eg access‑controlled Sharepoint sites.
Information should only be held in encrypted format on approved secure systems.
Only via University‑approved encrypted transfer methods; strictly limited recipients
Researchers should consider employing tools or adjusting settings so that the AI used will not retain user inputs or employ them to train AI.
When using AI in this way, researchers are advised to ensure that AI tools are set not to use inputs to train their models or store user input.
You should not therefore use such devices to access, transmit or store protected information.
Do not store protected information in personal cloud services such as Dropbox unless securely encrypted first.
- Roehampton has an institutional AI policy for research and knowledge exchange approved by Senate and reviewed through university research committees
- Its research governance framework also includes the Research Integrity and Ethics Committee, which oversees ethics approval and responds to external developments, but the provided sources do not set out a broader university-wide AI strategy or roadmap beyond the research domain
Owner: Edward Vallance, Dean of Research and Doctoral Study
Reviewed by: Research and Knowledge Exchange Committee, Research Degrees Committee Senate
Approved by: Senate
Review due date: September 2028
Update history: Senate, September 2025
Consider and respond to external developments, including legislation, regulations, professional body guidance, funder regulations, and sector standards, and make recommendations as appropriate
Oversee, and report on, the Ethics approval process, including scrutiny and approval or otherwise of high-risk Ethics applications
The Research Integrity and Ethics Committee meets twice a semester.
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 Roehampton University's AI Policies
Roehampton University has defined AI policies in 10 of 12 categories, with an overall coverage score of 83%.
Disclosure of permitted AI use is mandatory for both student assessed work and research outputs. Students must fully cite AI use when it is permitted in an assessment, while researchers must clearly acknowledge AI use, keep records of prompts and responses, identify the tool and date used, and provide acknowledgment or methodological explanation where appropriate.
Undisclosed or improper AI use is enforceable as academic misconduct. For taught students, misuse of AI carries sanctions up to permanent expulsion depending on level and prior offences; for postgraduate research students, misuse of AI can lead to permanent expulsion or required remediation. Research students submitting theses must use Turnitin, include an originality report, and work with supervisors if the similarity percentage is above 20%.
Roehampton requires university information to be classified and handled according to its classification level, which affects what may be entered into or stored on external tools. Internal-use information must stay on university-managed systems, restricted and confidential information must only be held on approved secure systems, and confidential information must use university-approved encrypted transfer methods; researchers are also advised to use AI tools that do not retain or train on user inputs, and staff/students must not store protected information in personal cloud services unless securely encrypted first.