Teaching & Learning
- Use of generative AI in coursework is permitted or prohibited depending on the assessment design under the university’s two-lane approach
- Some assignments allow AI use, but students must follow the specific lane requirements and any school or unit rules; using AI where it is not permitted or outside stated parameters is treated as poor academic practice or academic misconduct
Our two-lane approach means all coursework and assessments fit into one of two categories. Either students can use GenAI and submit the output in support of their own work, or students can’t use GenAI and must provide evidence of the process they followed to produce the work.
The use of GenAI for all modes of assessment should be considered carefully. This will usually entail one of the following approaches:
AI use not permitted
AI use expected, with no need to acknowledge
AI use expected, with some need to acknowledge
AI use permitted, but students can choose whether or not to use AI, and should acknowledge any use.
Using GenAI where not permitted, or failing to disclose use where use/disclosure is required, may be considered poor academic practice or academic misconduct.
Please note, some courses, programmes or schools may have additional guidance related to AI use in the curriculum, for assignments, and in exams and assessments. This is because there are disciplinary differences in what is considered best practice and acceptable use of AI technologies.
- Assessment use is governed by the same two-lane model: some assessments prohibit AI, while others require or permit it with varying disclosure rules
- The university also states that some courses or schools may issue additional exam and assessment guidance, and using AI where not permitted can lead to poor academic practice or academic misconduct findings
Our two-lane approach means all coursework and assessments fit into one of two categories. Either students can use GenAI and submit the output in support of their own work, or students can’t use GenAI and must provide evidence of the process they followed to produce the work.
The use of GenAI for all modes of assessment should be considered carefully. This will usually entail one of the following approaches:
AI use not permitted
AI use expected, with no need to acknowledge
AI use expected, with some need to acknowledge
AI use permitted, but students can choose whether or not to use AI, and should acknowledge any use.
Please note, some courses, programmes or schools may have additional guidance related to AI use in the curriculum, for assignments, and in exams and assessments. This is because there are disciplinary differences in what is considered best practice and acceptable use of AI technologies.
Using GenAI where not permitted, or failing to disclose use where use/disclosure is required, may be considered poor academic practice or academic misconduct.
- Students are told to use these tools critically because outputs can be inaccurate, biased, or fabricated, and they should not rely on them as authoritative sources
- The university encourages students to use GenAI to support learning and study, including explaining concepts, generating examples, and creating quizzes or study plans
Students can use GenAI in a variety of ways to support their learning, such as:
To understand or get a simplified explanation of a difficult concept, idea, or process.
To ask for examples that can help illustrate a theory or approach.
To make connections between ideas and think about a topic from a new perspective.
To create quizzes or flashcards to test understanding.
To suggest plans or schedules for revision and study.
GenAI can make mistakes. It can give inaccurate information, present opinions as facts, or invent references, quotations, data and events. This is sometimes called “hallucination”.
It is important to question and evaluate any AI-generated content before relying on it. Students should use their own judgement and verify key information using reliable academic sources.
- Where programming work is assessed, whether AI coding assistance is allowed depends on the assessment’s lane and any local course or school guidance
- The university does not set a separate institution-wide rule specifically for code-generation in programming assignments beyond the general two-lane assessment framework
Our two-lane approach means all coursework and assessments fit into one of two categories. Either students can use GenAI and submit the output in support of their own work, or students can’t use GenAI and must provide evidence of the process they followed to produce the work.
Please note, some courses, programmes or schools may have additional guidance related to AI use in the curriculum, for assignments, and in exams and assessments. This is because there are disciplinary differences in what is considered best practice and acceptable use of AI technologies.
Research
- Research outputs must not list AI tools as authors, and researchers remain responsible for the accuracy and integrity of any content produced with AI assistance
- The university also requires researchers to check funder, publisher, and disciplinary expectations before using AI in research writing or manuscript preparation
Authors are responsible for the accuracy, integrity and originality of their work, even if they have used GenAI tools to help draft or edit text.
Generative AI tools cannot be listed as authors on publications.
Researchers should check and follow any guidance from publishers, funders, or disciplinary bodies on the use of GenAI in research and scholarly communication.
- Researchers may use AI tools in research activity only with caution and must protect confidentiality, personal data, and contractual or legal restrictions
- The university warns that uploading unpublished or sensitive research material into public GenAI tools can create security, privacy, and intellectual property risks
Researchers must not upload confidential, sensitive, personal, or commercially restricted information into public GenAI tools unless explicitly authorised and compliant with data protection, confidentiality and contractual obligations.
The use of GenAI may pose risks to research integrity, data security, confidentiality, intellectual property, and compliance with legal and ethical standards.
Particular care should be taken when using GenAI tools in relation to unpublished research data, participant information, or material subject to non-disclosure agreements or intellectual property protection.
- The university explicitly connects AI use in research to research integrity obligations
- Researchers are responsible for maintaining accuracy, originality, confidentiality, and compliance with legal and ethical standards, and must not use public GenAI tools with restricted material unless authorised and compliant
The use of GenAI may pose risks to research integrity, data security, confidentiality, intellectual property, and compliance with legal and ethical standards.
Authors are responsible for the accuracy, integrity and originality of their work, even if they have used GenAI tools to help draft or edit text.
Researchers must not upload confidential, sensitive, personal, or commercially restricted information into public GenAI tools unless explicitly authorised and compliant with data protection, confidentiality and contractual obligations.
Academic Integrity
- Disclosure requirements depend on the assessment lane
- The university also provides general guidance on citing and referencing sources
- In some assessments no acknowledgement is needed, in others acknowledgement is required, and students are directed to follow the specific instructions provided; undisclosed AI use where disclosure is required can trigger poor academic practice or academic misconduct
The use of GenAI for all modes of assessment should be considered carefully. This will usually entail one of the following approaches:
AI use not permitted
AI use expected, with no need to acknowledge
AI use expected, with some need to acknowledge
AI use permitted, but students can choose whether or not to use AI, and should acknowledge any use.
Using GenAI where not permitted, or failing to disclose use where use/disclosure is required, may be considered poor academic practice or academic misconduct.
In all cases, the task or assessment should clearly state whether or not students are permitted to use GenAI tools, and if they are, whether and how that use should be acknowledged.
- Undisclosed or prohibited AI use may be handled under the university’s poor academic practice or academic misconduct processes
- The university’s guidance focuses on misconduct procedures and does not define a specific institution-wide AI detection-tool rule in the cited materials
Using GenAI where not permitted, or failing to disclose use where use/disclosure is required, may be considered poor academic practice or academic misconduct.
Academic misconduct is action or attempted action that may result in creating an unfair academic advantage for yourself or an unfair academic disadvantage for another member of the academic community.
Academic misconduct includes, but is not limited to:
cheating in an examination
aiding someone else to cheat in an examination
plagiarism
self-plagiarism
fabrication or falsification of data or signature
contract cheating or commissioning.
Institutional & Administrative
- Staff may use GenAI in education-related work, but they must do so ethically and lawfully, with human oversight and professional responsibility retained by the staff member
- The university warns staff not to input personal, confidential, or commercially sensitive information into unapproved tools and says outputs should be reviewed for accuracy, bias, and appropriateness before use
Staff remain responsible for any content produced with the support of GenAI tools and should exercise professional judgement when using such tools in teaching, assessment, feedback, and administration.
You should always review and, where necessary, edit AI-generated outputs before using them. This is particularly important where outputs may affect students, colleagues, or external stakeholders.
Do not enter confidential, personal, commercially sensitive or otherwise restricted information into GenAI tools unless the tool has been approved for that use and appropriate safeguards are in place.
- The university restricts what information can be entered into GenAI tools and requires compliance with information security, data protection, and acceptable use rules
- It promotes approved institutional tools such as Microsoft Copilot and states that confidential, personal, commercially sensitive, or otherwise restricted data must not be entered into AI tools unless the tool is approved and safeguards are in place
Do not enter confidential, personal, commercially sensitive or otherwise restricted information into GenAI tools unless the tool has been approved for that use and appropriate safeguards are in place.
Microsoft Copilot is the University’s recommended generative AI tool for work-related use because it offers commercial data protection when used with your University account.
University information must be handled in accordance with its classification and with appropriate security controls.
You must not use University IT facilities in a way that breaches information security, data protection or confidentiality obligations.
- The university has adopted institutional principles for GenAI in education and implemented a university-wide two-lane framework for assessment
- It also describes an organised governance and leadership response, including a GenAI Academic Lead for Education and GEOG as part of the university’s response to generative AI in education
The University of Bath has agreed a set of principles for the use of generative artificial intelligence (GenAI) in education.
Our two-lane approach means all coursework and assessments fit into one of two categories.
announcing James Fern as GenAI Academic Lead for Education
Introducing GEOG: leading the University of Bath’s response to generative AI in education
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 Bath's AI Policies
University of Bath has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
Disclosure requirements depend on the assessment lane. In some assessments no acknowledgement is needed, in others acknowledgement is required, and students are directed to follow the specific instructions provided; undisclosed AI use where disclosure is required can trigger poor academic practice or academic misconduct. The university also provides general guidance on citing and referencing sources.
Undisclosed or prohibited AI use may be handled under the university’s poor academic practice or academic misconduct processes. The university’s guidance focuses on misconduct procedures and does not define a specific institution-wide AI detection-tool rule in the cited materials.
The university restricts what information can be entered into GenAI tools and requires compliance with information security, data protection, and acceptable use rules. It promotes approved institutional tools such as Microsoft Copilot and states that confidential, personal, commercially sensitive, or otherwise restricted data must not be entered into AI tools unless the tool is approved and safeguards are in place.