University of Surrey

4.4. Considered – avoid inputting sensitive data into unapproved tools; assess bias and accuracy risks. 4.4.1. Protect Information: To reduce the risk of data breaches and misuse, personal, confidential, and commercially sensitive information (including Intellectual Property) must NOT be entered into GenAI tools—internal or external— unless they are approved by IT Services and meet the organisation’s security and compliance requirements. 4.4.2. Mindful input: Being cautious about what you input into GenAI tools is crucial because GenAI generates responses based on its training data, which may not always be accurate or appropriate. 4.4.3. Bias awareness: Recognising and mitigating biases in GenAI models helps prevent discrimination and ensures fair outcomes. 4.5. Transparent – Clearly label and reference GenAI-generated content where applicable. 4.5.1. Transparency: About the use of GenAI fosters trust and accountability. 4.5.2. Referencing GenAI: Clearly labelling GenAI-generated content helps users understand the source and context of the information.

Collected: 2026-10-08 · 97–113
SHA-256: 83a8b85cab9a091749c6da41671e8945b95df4a37123120ddd88747ca713695d

2. Scope and Exceptions to the Procedure 2.1. This applies to: 2.1.1. All University of Surrey employees, subsidiaries, contractors, visiting researchers and temporary employees 2.1.2. All GenAI use for university related work, teaching, research and administration 2.1.3. The design, development, or deployment of GenAI enabled products or services within the University context 2.2. This is the overarching procedure on GenAI at the University. Other AI related guidelines, support information, training material are produced from time to time and are available on the AI page on SURREYNET Artificial intelligence at Surrey | Surreynet

Collected: 2026-10-08 · 54–65
SHA-256: 83a8b85cab9a091749c6da41671e8945b95df4a37123120ddd88747ca713695d

(vi) passing off work as original that may have been generated by Artificial Intelligence. This could be, but is not limited to, text, images, music or code. Other forms of academic misconduct 20. Other forms of academic misconduct include, but are not limited to: (i) failure to declare third party assistance, including from generative Artificial Intelligence, in the presentation of assessed work (other than assistance by a department of the University), including language, syntax, spelling and layout or failure to provide the draft material submitted to the third party to proof-read and/or correct. Guidance on the appropriate use of generative Artificial Intelligence for specific assessments will be communicated by the Module Team; (ii) assistance in the completion of assessed work from third parties, including from generative Artificial Intelligence, through proof-reading and correcting English or a target language (not including use by the student of dictionaries, thesauruses and spell-checkers) where the learning outcomes for a module include a specific requirement to demonstrate facility with written English or a target language; (iii) fabricating results from laboratory or other work or misrepresenting data; (iv) introducing unauthorised textual materials into an assessment venue or ancillary area such as cloakroom or toilets. This can include not only notes clearly relevant to the assessment but any form of written material, either on paper or on a student’s body; (v) having an unauthorised mechanical or electronic device on one’s person within an assessment venue or ancillary area, such as a cloakroom or toilets. Where such devices, including mobile phones and smart watches, are permitted to be brought into an assessment venue, they must be switched off and placed in the container provided for this purpose. The container must then be placed on the floor in full view of the invigilators before the start of the assessment and for the duration of the assessment; (vi) passing off the work of others as your own during an online assessment, including content and/or writing produced by generative Artificial Intelligence. (vii) accessing unauthorised online material/web pages during an online assessment.

Collected: 2026-10-08 · 345–374
SHA-256: b642850f027e275614aa23d34bf840cee3f0ceba161ca4cd77dc506da07331f2

35. The student receives five University working days’ notice of the formal discussion and is invited to submit any representations in relation to the concerns, including evidence of special circumstances. The student is also informed of the support provided by the Surrey Students’ Union and that they can be accompanied by a friend or an official of the Surrey Students’ Union at the discussion. The letter outlines the nature of the allegation and the potential outcomes of the formal discussion. A copy of all relevant material is also provided to the student in advance of the meeting. In exceptional circumstances and with written agreement of the student, the formal discussion may take place earlier than five University working days. 36. The formal discussion is also attended by the Module Leader or, in the case of dissertations, by the first marker. If these persons are unavailable, they can nominate another member of the module team to attend. The student and the AIOs will receive details of the allegation and the supporting evidence (e.g. a Turnitin®/Authorship Investigate report), where relevant. AIOs will not be informed of any previous findings of either poor academic practice or academic misconduct until after they have made their decision and, if relevant, are determining the penalty. The only other exception is when the student misleads the AIOs in relation to any previous findings. If this occurs, the Academic Administration team will correct any inaccurate information. 37. Where a student is unable to attend the meeting on the specified date and time, they can request an alternative provided there are valid reasons supported by appropriate evidence. The formal discussion can take place in person or online or by telephone. 5 The AIOs should not have a close relationship with the student (or students) and should not have been involved in previously proven misconduct allegations against the student. 9 38. The formal discussion provides an opportunity for the student to explain how they approached the assessment task and for the student to be shown how the suspected academic misconduct has been identified. In cases of suspected collusion, consideration will be given as to whether the assessment instructions were sufficiently clear as to whether the work was to be undertaken singly or collectively. In cases of possible academic misconduct in group work, consideration will be given as to whether the group work was designed to produce a single piece of work with a mark in common for all members of the group or whether discrete elements were produced by individual members. 39. The student is also given the opportunity to raise any previously undisclosed special circumstances. 40. Should a student choose to waive their right for a formal discussion to take place, they must notify the University of their decision in writing (via email). If the student does not attend the formal discussion and there are no valid reasons for non-attendance, the AIOs will reach a finding based on the information available to them.

Collected: 2026-10-08 · 565–607
SHA-256: b642850f027e275614aa23d34bf840cee3f0ceba161ca4cd77dc506da07331f2

Penalties for proven cases of academic misconduct Table 1: Penalties that an Academic Integrity Officer, the Academic Administration team and an Academic Misconduct Panel instructs a Board of Examiners to apply 1213 Instances of Scale of penalties where academic Scale of penalties where academic academic misconduct is found to have taken misconduct is found to have taken misconduct place during a first assessment attempt place during a reassessment attempt First instance Standard cases: Standard cases: Mark of zero for the unit of assessment. Mark of zero for the unit of assessment. If the module is failed – reassessment is If the module is failed – no further allowed. Reassessment penalty is applied reassessment is allowed unless the to the re-assessed unit(s) of assessment. student is entitled to a third attempt. Regulations for taught programmes apply. Serious cases: Mark of zero for the unit of assessment Serious cases: 12 Decisions as to whether misconduct should be categorised as standard, serious, or severe are made at the sole discretion of the AIO(s), AMP or Academic Administration staff members. They will consider the facts of the case alongside relevant, contextual evidence and make a determination on the balance of probabilities. 13 Findings of misconduct made within the previous six calendar years, including during separate registration periods, will normally be considered when penalties are applied. 16 AND Mark of zero for the unit of assessment If the module is passed overall – the AND module mark is capped at the pass mark. If the module is passed overall – the module mark is capped at the pass mark. If the module is failed – reassessment is allowed. Reassessment penalty is applied to the re-assessed unit(s) of assessment. If the module is failed – no further In cases, where following the reassessment is allowed unless the reassessment attempt, the module is student is entitled to a third attempt. passed, the module mark is capped at the Regulations for taught programmes pass mark. apply. Second instance Mark of zero for the unit of assessment Mark of zero for the unit of assessment AND AND If the module is passed overall – the If the module is passed overall – the module mark is capped at zero and module mark is capped at zero and module credits are awarded (for credit module credits are awarded (for credit bearing modules). bearing modules). If the module is failed – reassessment is If the module is failed – no further allowed. Reassessment penalty is applied reassessment is allowed unless the to the re-assessed unit(s) of assessment. student is entitled to a third attempt. In cases, where following the The module mark is capped at zero. reassessment attempt, the module is No module credits are awarded. passed, the module mark is capped at Regulations for taught programmes zero and module credits are awarded (for apply. credit bearing modules). Third instances and Termination of registration. Regulations Termination of registration. severe cases for taught programmes apply Regulations for taught programmes apply

Collected: 2026-10-08 · 946–1016
SHA-256: b642850f027e275614aa23d34bf840cee3f0ceba161ca4cd77dc506da07331f2

Review by an OSCAR Case Manager 17. A student wishing to appeal against the findings or outcome of an academic misconduct process, is required to do so within 10 University working days of being notified of the decision. Appeals should be submitted to OSCAR on-line and in accordance with published requirements as to the format, content and length of submission. Further details are available on the OSCAR web pages. If the appeal is received on time, it will be assigned to a Case Manager. If a student is unable to provide all supporting evidence by the appeal submission deadline (for example, if they are awaiting receipt of a medical letter), then they are still required to submit their appeal within the 10 University working day timeframe (including a complete supporting statement and any available evidence) but may request an extension in respect of the submission of any pending evidence. Such requests should explain the reason the extension is required and be made by email to OSCAR at the time of the appeal submission. 18. If an appeal is received after the 10 University working days deadline the student will be asked to provide reasons and accompanying independent supporting evidence as to why the appeal is late. The reasons and supporting evidence will be considered by an OSCAR Case Manager to determine whether the reasons are valid. Consultation may take place with other members of staff in complex or unusual cases, and this will be recorded. A decision on the validity of the reasons will normally be made within 20 University working days of receiving the final submission of information from the student. If the reasons are not deemed valid, the appeal will not be considered and the student will be offered a Completion of Procedures letter. 1 If there are valid reasons, the appeal will be accepted and assigned to a Case Manager.

Collected: 2026-10-08 · 170–193
SHA-256: b491302e1dffdab6a741b842aef088a471f7b6d3345f20842fb766211919cab0

Grounds for making an appeal 9. When making an appeal the student is required to show that they have evidence to demonstrate that one or more of the following grounds apply: • that staff or bodies have failed to follow regulations and/or procedures or have failed to follow them with due care • that staff or bodies have shown bias or prejudice towards the student in the way they have made the relevant decision • that relevant new evidence has become available that should be considered and there are valid reasons why it was not provided earlier • that the decision was unreasonable and/or the penalty imposed was not proportionate in all of the circumstances 10. Where new evidence is raised, further investigations may need to be carried out which may require the appeal to be delayed pending the outcome of these investigations and to give the student an opportunity to consider any new information obtained. Circumstances in which the University will not consider an appeal 11. An appeal against a decision where only the opinion of academic expert(s) will suffice is deemed to be a challenge to academic judgement and will not be considered.

Collected: 2026-10-08 · 112–128
SHA-256: b491302e1dffdab6a741b842aef088a471f7b6d3345f20842fb766211919cab0

My AI Surrey, that provides equitable access to major Large Language Models (LLMs), like ChatGPT, for all Surrey students. My AI Surrey will help to empower you to harness AI, learning to use the tools effectively and graduate with the capabilities for AI-enabled problem solving that the employment market demands. My AI Surrey is a university-supported space enabling you to explore and use AI with confidence by providing you with access to Large Language Models (LLMs) inside a secure Surrey environment. My AI Surrey is still in development. We recognise this is a rapidly evolving space and more and different tools will be coming over the next few weeks and months, but we wanted to provide equitable access to major LLMs as soon as possible. Visit My AI Surrey Using AI in your studies Generative AI tools may be used to support your studies, but they may only be used to support, not replace, your authorship. All submitted work must be your own intellectual output. It is important to remember that: AI-generated information may be inaccurate or fabricated (including false references and DOIs); all outputs must be critically evaluated and verified against authoritative academic sources; presenting AI-generated content as original work, without acknowledgment, may constitute academic misconduct. In summary, while AI tools may help you, you must always remain in control of your submitted work.

Collected: 2026-10-08 · 201–213
SHA-256: d48cafb4a33e2f3862e520ef8088761a05fdf4ae2aef80a88f10ad49091cd172

Reasonable adjustments 4. Reasonable adjustments to the processes within these Regulations, including the extending of deadlines for student responses, will be made upon the production by the student of relevant third party evidence which demonstrates the need for those adjustments. Exceptional circumstances 5. In exceptional circumstances it may be appropriate to amend the procedures set out in these Regulations, for example, where strict application of the Regulations would result in substantial unfairness to a student or a student is in some way at risk because of health or disability. Such cases will be rare and each will be treated on their own merits. Burden of proof 6. When making an appeal it is for the student to show that one or more of the grounds listed in regulation 9 below apply. Standard of proof 7. The standard of proof is that of the balance of probability; that it is more likely than not that something was or was not the case.

Collected: 2026-10-08 · 74–90
SHA-256: b491302e1dffdab6a741b842aef088a471f7b6d3345f20842fb766211919cab0