University of Bristol

We consider the use of AI or translation tools to be cheating if used for more than: generating the occasional short phrase within a sentence, checking basic grammar and spelling. Some assessments may allow more comprehensive use of these tools, but this will be detailed in your assessment instructions.

10.1. Cheating is where a student is considered to have broken the rules of an assessment or sought to gain an unfair advantage through unauthorised means. A student engaging in any of the following is considered to be cheating: h. unauthorised use of artificial intelligence to complete a piece of work (that means a student does not demonstrate the intended learning outcomes for the relevant unit through that assessment).

14.20. A student may appeal a decision made by a panel to apply a mark penalty for academic misconduct as set out in section 15 below. Appeals against academic decisions of an exam board are covered by the regulations in the Taught Code on Student Outcomes Appeals and the Student Outcomes Appeal Procedure. 15.2 To submit an appeal, the student must complete and submit an academic misconduct appeal form within 14 days of the student being sent the decision of the panel. 15.3 A day refers to a calendar day in which the University is generally open and so includes weekends but does not include public holidays in England or University closure days as marked on the University’s website, unless otherwise stated.

Owner Academic Quality and Policy Office Approved by University Academic Quality Committee Version 2.0 Approval date 10 June 2026 Effective date 01 August 2026 (for the 2026/27 academic year)

6.2. Software may be used as to help identify potential academic misconduct. Any suspected cases are reviewed before any action is taken. Where there are any contractual, security or safety obligations to safeguard student work, a request for exemption from using third-party software to check for plagiarism may be made by a student or appropriate member of staff (e.g. dissertation supervisor). The request is considered for approval by the relevant Faculty Associate Pro Vice-Chancellor Education and Students or delegate. If the request is approved, a relevant member of staff instead undertakes a manual check on the dissertation for academic integrity and plagiarism.

14.12. Once the interview is complete, the panel determines if, on the balance of probabilities, misconduct has occurred or not. The academic misconduct panel may decide to defer a decision until further investigation has taken place. In such cases the panel decides whether they need to engage with the student again to discuss any new material that has been considered as part of the decision-making process.

14.6. The student is notified in writing of the suspected case (including the pieces of work affected) and invited to interview with a panel. The interview can be in-person or online. 14.7. The student can request: the option of answering written questions instead of, or as well as, attendance at the panel; and, to see the evidence considered by the panel. 14.9. The student may elect not to attend an interview and to submit a statement to the panel instead. Where the student does not attend an interview, unless prevented by unforeseeable circumstances, the panel proceeds to reach its conclusions without the student’s attendance.

14.8. The student may be accompanied by another person at the panel interview. This person may address the panel and confer with the student during the meeting, however, they are not permitted to answer any questions on behalf of the student, unless in exceptional circumstances where they are a qualified lawyer, and the student is unable to do so themselves.

14.15. In cases where there is an exceptionally high level of dishonesty and seriousness of the misconduct such that the student's status at the University is called into question, the panel may refer a case to the Student Disciplinary procedure. 14.16. The student is informed of the panel’s decision within two weeks of the date of the panel.

15.1. Students may appeal the decision of an Academic Misconduct panel on one or more of the following grounds: New supporting evidence, relevant to the finding of academic misconduct, is available that could make a difference to the outcome and which the student was unable, for good reason, to disclose at the time of the panel. Evidence relating solely to exceptional circumstances will not constitute grounds of appeal unless in exceptionally severe cases (see 22.17 in the Taught Code for details). There is evidence of significant procedural error which warrants consideration. Disagreement with the academic judgment of the panel will not constitute a valid ground for appeal.

15.6 In exceptional circumstances, the University may consider a request to submit an appeal outside of published time limits. Students must provide a clear explanation and supporting evidence demonstrating why they were unable to meet the deadline. If an appeal is not submitted within the permitted time limit, and no new deadline has been agreed, the appeal will be closed and no further consideration will be given.

15.9 Disabled students with a condition, impairment or learning difference covered by a Study Support Plan (SSP) are not required to submit additional medical or supporting evidence to re-establish the existence of their disability when submitting an appeal. The SSP constitutes the established evidence base for their disability. However, where the appeal is based on a worsening or acute episode of their condition, students are expected to provide evidence of the impact and timing of the change in circumstances.

15.14 For all outcomes, the decision will be communicated within 14 days of the completed appeal form and all relevant evidence being submitted. 15.15 The University endeavours to respond within the timeframes that are outlined. There will occasionally be circumstances when, for good reason, the University will need to extend the timeframe. When this occurs, the University will inform the student and provide regular progress updates.

16.1. The OIA provides an independent scheme for the review of student complaints. The OIA will only consider cases when the University’s internal procedures have been exhausted. It will not intervene in matters which turn purely on academic judgment. 16.3. If an appeal is upheld (fully or partially), a Completion of Procedures letter is not issued automatically. Students can request a letter within one month of the outcome if required.

13.3. Any student referred to an AIAC can appeal this decision. The student may raise their appeal by submitting the AIAC Appeal Form by email to the Student Resolution Service (SRS) within 21 days of being informed they are required to attend an AIAC. The student must set out in writing the grounds on which the appeal is based, as set out in regulation. 13.4. The appeal is heard by a Faculty Associate Pro Vice-Chancellor Education and Students, or equivalently experienced academic who has not been involved in the case.

12.1. An undergraduate student is invited to attend an Academic Integrity Awareness Course if: it is their first offence, and they are not in their final year of study, and the case is not considered too serious that it cannot be addressed by additional training. A taught postgraduate student may be invited to either attend an Academic Integrity Awareness Course or have a supportive discussion with the unit director and redo the affected work, subject to stated conditions.

For example, transgressions relating to research by a student may be triggered via the research misconduct policy or the academic regulations. The appropriate route should be decided upon by the Faculty Academic Director (PGR) or the Associate PVC (PGR) based on both the potential offence and the level of study.

Scope - This document applies to: All students registered on programmes of study leading to an award of the University of Bristol at undergraduate, postgraduate taught and postgraduate research levels. Applies to academic year: 2026/27 Version 2.3 Date current version approved 10/08/2026 Date current version published 17/08/2026 Superseded documents From version 2.2 minor editing changes

4.1.3. Misuse of Artificial Intelligence (AI): Using AI in a manner that is not authorised for an assessment. 4.1.8. Breaching ethical or other obligations: Failing to comply with the University’s Ethics of Research Policy or with any regulatory, legal, professional or contractual obligations, including on copyright and Intellectual Property.

4.3.3. Serious academic misconduct – Serious instances of misconduct where the student would gain a substantial unfair advantage or where there is an intention to deceive, e.g. extensive misuse of Artificial Intelligence in an unauthorised manner.

If you or your supervisor has any questions about the acceptable use of AI tools in your studies, please contact your School PGR Director. And if you are worried that you may have used AI tools inappropriately in your thesis writing so far, your School PGR Director can give you advice on how to address this.

Review the use of Turnitin AI writing detection within their area of responsibility and decide whether to persist with its use until the next university checkpoint at the end of 24/25.

22.6. The AIAC is available to undergraduate students (excluding those in their final year of study but including those on Foundation programmes) and taught postgraduate students where: this is their first offence, and the offence is not sufficiently serious that it cannot be addressed by training. 22.7. Full attendance and engagement is required for successful completion of the training. If a student does not attend, or attends but refuses to engage or is disruptive, the relevant assessment is awarded a mark of zero. 22.8. Any student referred to an AIAC can appeal this decision by submitting the AIAC appeal form within 21 days of being informed they are required to attend the course.

Or, if the panel judges that academic misconduct has occurred, it determines the extent to which the assessment can still evidence the student’s learning and the level of advantage they may have gained, and either: Notes the case on the student’s record for future reference with no penalty applied; Applies one of the following penalties to the original piece of work by awarding a lower mark than would otherwise have been awarded by: capping the mark for the assessment at the pass mark; applying a mark reduction of 10 marks; applying a mark reduction of 20 marks; awarding a mark of zero for the assessment; in assessments that are marked on a pass/fail marking scheme, awarding a ‘fail’ mark for the assessment.

You have 21 days (including weekends) to appeal from the day you are contacted about attending an AIAC. This does not include English public holidays or University closure days. To appeal, you must: Download the Academic Integrity Awareness Course appeal form. Complete all sections. Submit the form to [email protected]. You must still attend the AIAC while your appeal is under review.

The University of Bristol has enabled Turnitin’s AI detection tool to support academic integrity. However, staff are advised to use the tool only when concerns arise, as academic integrity is best supported through effective assessment design. This score is an indicator rather than evidence of AI use. It should not be interpreted as proof that AI has been used and must always be considered alongside assignment settings, academic judgement and the wider context of the assessment. An AI writing score is one source of information that instructors may consider when reviewing a student's work. However, on its own, it is not evidence of academic misconduct.

Human academic judgement remains central. Teaching and assessment should develop and foreground academic and critical skills so that students can make responsible decisions about their academic work. Decisions about teaching, assessment and academic standards rest with academic staff and exam boards.

Some tasks may be designed to use AI as part of your assessment or studying. If you follow exactly what has been requested by your school, this will not be considered contract cheating.

This guidance sets out the University of Bristol's approach to using generative AI on taught degree programmes. The responsible use of generative AI is foundational to our educational vision to inspire students, and to equip well-rounded graduates with skills for the future. Generative AI should not replace activities that develop intellectual rigour, student agency, and students' capacity to work through complex problems themselves.

Category 1: prohibited (not allowed) - You cannot use AI tools at all, including spelling and grammar checkers. Category 2: minimal (a little) - You can use AI tools for spelling and grammar check. You can use it for rewording occasional phrases, but not sentences, large parts of text or paragraphs. Category 3: selective (certain tasks) - AI tools can be used in certain ways. Your lecturer will tell you what you can use it for. Category 4: integral (important) - AI tools are needed for the assessment.

This guidance sets out the University of Bristol’s approach to using generative AI on taught degree programmes.

Recommend that students use Microsoft Co-pilot through the Edge browser as the default, universally available generative AI tool, where one is needed. All staff are asked not to: Upload student work to externally-hosted AI tools, unless this tool has been authorised for use by IT Services. Use AI to mark or generate staff feedback on student work, unless this has been discussed in advance with your SED/FED and the AI in education team at [email protected] who may enable you to undertake an authorised pilot.

You should not use AI tools to mark or provide feedback on student work without discussing your plans with SEDs/FEDs and our central AI team. If agreed, you will need to discuss, explain, and agree any use of generative AI for feedback with students, in line with our principles of transparency and dialogue. You should not upload student work to externally-hosted generative AI tools, unless this tool has been authorised for use by IT Services.

Translation tools may only be used for assessments if the assessment brief specifically allows it (for example, annotating or improving on an AI translation). Do not upload university-sensitive data into AI tools unless you have been given permission by academic staff. This includes: intellectual property; personal information; research data. This list is not exhaustive. If you are not sure, speak to your lecturer or school.

Recommend that students use Microsoft Co-pilot through the Edge browser as the default, universally available generative AI tool, where one is needed. All staff are asked not to: Upload student work to externally-hosted AI tools, unless this tool has been authorised for use by IT Services. Use AI to mark or generate staff feedback on student work, unless this has been discussed in advance with your SED/FED and the AI in education team at [email protected] who may enable you to undertake an authorised pilot.

Microsoft Copilot for Edge browser is available to all staff and students free of charge through the University Microsoft licence. When using Copilot for Edge, you and your students should log in with your university email. This provides you with commercial data protection, meaning your chat history is not saved, and the data is not used to train AIs, ensuring that university inputs remain secure.