Teaching & Learning
- Undisclosed or unauthorized AI use in assessed work is treated as academic misconduct
- Where it is allowed, students remain responsible for the accuracy, integrity, and proper acknowledgement of any AI-assisted work
- For assessed coursework, use of generative AI is only permitted when a teacher has explicitly allowed it for a specific assignment
There may be some uses of AI tools that are acceptable in some circumstances but this should be explicitly set out by your teacher for each assignment.
You should not use AI tools for assessed coursework unless specifically instructed or allowed by your teachers. If in doubt, please ask.
Any use of generative AI tools in relation to an assessment should always be acknowledged and referenced appropriately.
Remember, output from AI tools can be incorrect or fabricated. You remain responsible for checking the accuracy of any information generated, and are accountable for any work submitted for assessment.
Using an AI text generation tool such as ChatGPT to produce work for assessment where this is not explicitly permitted by the teacher
Submitting work generated in whole or in part by artificial intelligence as one’s own work, where the use of artificial intelligence has not been expressly permitted by the teacher
- For assessments broadly, including formal submitted assessments, AI use is not allowed unless the teacher has expressly permitted it
- The university treats using AI-generated material in assessment without permission as misconduct, and where AI use is allowed it must be acknowledged
You should not use AI tools for assessed coursework unless specifically instructed or allowed by your teachers. If in doubt, please ask.
Any use of generative AI tools in relation to an assessment should always be acknowledged and referenced appropriately.
Using an AI text generation tool such as ChatGPT to produce work for assessment where this is not explicitly permitted by the teacher
Submitting work generated in whole or in part by artificial intelligence as one’s own work, where the use of artificial intelligence has not been expressly permitted by the teacher
- SOAS permits students to use generative AI as a study aid for personal learning, such as clarifying concepts, summarising, translation support, brainstorming, and practice-style help
- The guidance warns students to use these tools critically because outputs may be inaccurate, biased, or fabricated, and students must still develop their own understanding and skills
There are many possible uses for such tools in relation to learning, including:
• Getting explanations or summaries of difficult concepts or texts.
• Translation support.
• Generating ideas for topics, arguments and structures.
• Study support such as creating quiz questions, revision notes or flashcards.
• Practising language skills.
Generative AI tools can be useful and save time, but there are also significant drawbacks. They can produce inaccurate information, biased content, and fabricated references. Their outputs can appear plausible even when wrong. You should use them critically and carefully.
These technologies should support, not replace, your own learning and development. You still need to engage with course materials, think independently, and develop your own academic skills.
Research
- Separate university-wide research-manuscript rules for staff or non-assessed scholarly publications are not defined in the provided sources
- SOAS applies its student AI guidance to assessed research writing such as dissertations and similar submissions: AI use is not allowed unless expressly permitted by the teacher, and any permitted use must be acknowledged
You should not use AI tools for assessed coursework unless specifically instructed or allowed by your teachers. If in doubt, please ask.
Any use of generative AI tools in relation to an assessment should always be acknowledged and referenced appropriately.
Submitting work generated in whole or in part by artificial intelligence as one’s own work, where the use of artificial intelligence has not been expressly permitted by the teacher
- The provided sources do not set out explicit rules on using AI for research data collection, analysis, statistical processing, synthetic data generation, or interpretation of results
not defined
- The provided sources do not explicitly define AI-specific rules for grant proposals, ethics applications, or research-integrity declarations
- They establish general research ethics and integrity expectations, but no explicit AI standard for these research governance documents is stated
not defined
Academic Integrity
- The university also instructs students to keep records of prompts and outputs and to cite AI use in line with the relevant referencing style
- SOAS requires students to acknowledge and reference any use of generative AI in assessments, and the guidance provides a model declaration for doing so
Any use of generative AI tools in relation to an assessment should always be acknowledged and referenced appropriately.
You should clearly state how you used the tool and for what purpose. For example:
I acknowledge the use of [name of AI tool] to [describe use, e.g. generate ideas / summarise a text / check grammar]. The output was reviewed and edited by me, and I take responsibility for the final submission.
You may also need to cite the use of AI according to the referencing style required by your department.
Keep records of any prompts used and the outputs generated, in case your teacher asks you to explain your process.
- SOAS classifies unauthorized AI-generated assessment work as academic misconduct and places it within formal misconduct procedures
- The provided sources specify the offense and indicate that penalties are handled under the academic misconduct policy, but they do not define a university position on AI detection tools
Using an AI text generation tool such as ChatGPT to produce work for assessment where this is not explicitly permitted by the teacher
Submitting work generated in whole or in part by artificial intelligence as one’s own work, where the use of artificial intelligence has not been expressly permitted by the teacher
Academic misconduct is any action or attempted action that may result in a student obtaining an unfair academic advantage in the assessment of their work, whether or not there is any intention to deceive.
Institutional & Administrative
- One student-facing statement indicates that teachers may decide assignment-specific permissions for student AI use
- The provided sources do not define explicit university-wide rules for faculty or staff use of AI in grading, feedback, lesson planning, recommendation letters, or administrative communications
There may be some uses of AI tools that are acceptable in some circumstances but this should be explicitly set out by your teacher for each assignment.
- The provided sources do not name a general approved AI platform list, but they clearly require approval before restricted data is entered into external AI tools
- Staff and students must use institution-approved IT services, protect personal data, and ensure processing complies with data protection and information security requirements
- SOAS prohibits putting confidential, personal, or other restricted university information into public AI systems unless the service has been approved and a compliant agreement is in place
Where information is confidential, personal, commercially sensitive, subject to legal privilege, export controls, contractual restrictions, or otherwise classified as restricted under SOAS information governance requirements, users must not input that information into publicly available generative AI systems unless the service has been approved by SOAS and appropriate contractual, security and data protection safeguards are in place.
Users must ensure that any processing of personal data using AI tools complies with UK data protection legislation and SOAS policies.
Use only IT services approved or provided by SOAS for institutional business, unless explicit authorisation has been granted.
Never disclose confidential information, personal data, or information subject to legal, regulatory, contractual, or ethical restrictions to unauthorised third parties or systems.
- The provided sources show university-level statements and policy controls, but they do not define a formal AI strategy, committee structure, or roadmap for adoption
- SOAS has issued institution-level guidance framing generative AI as a developing area that should be used cautiously, transparently, and within existing academic and governance rules
SOAS recognises that generative artificial intelligence (AI) technologies such as ChatGPT are rapidly developing and may have legitimate uses in support of learning and teaching.
The use of such tools raises important questions about academic integrity, assessment, transparency and the development of students’ own knowledge and skills.
Our approach at SOAS is to encourage critical engagement with these technologies while ensuring that academic standards and the value of our qualifications are protected.
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 SOAS, University of London's AI Policies
SOAS, University of London has defined AI policies in 11 of 12 categories, with an overall coverage score of 92%.
SOAS requires students to acknowledge and reference any use of generative AI in assessments, and the guidance provides a model declaration for doing so. The university also instructs students to keep records of prompts and outputs and to cite AI use in line with the relevant referencing style.
SOAS classifies unauthorized AI-generated assessment work as academic misconduct and places it within formal misconduct procedures. The provided sources specify the offense and indicate that penalties are handled under the academic misconduct policy, but they do not define a university position on AI detection tools.
SOAS prohibits putting confidential, personal, or other restricted university information into public AI systems unless the service has been approved and a compliant agreement is in place. Staff and students must use institution-approved IT services, protect personal data, and ensure processing complies with data protection and information security requirements. The provided sources do not name a general approved AI platform list, but they clearly require approval before restricted data is entered into external AI tools.