Generative AI, including Large Language Models (LLMs) like ChatGPT, is a rapidly evolving technology that can generate new text, images, video, audio, or code. These tools can be used to summarise long articles, answer specific questions, or even write code for a described function. However, their use comes with certain responsibilities and considerations.
This guidance applies to all staff at St George's, University of London. It is intended to help you understand how to use these tools effectively and responsibly in your work.
Summary of guidance
-
Never input sensitive information or personal data into these tools. Generative AI tools should not be used to process classified, sensitive, or personal data. Always adhere to the principles of GDPR and St George's data protection policy.
-
Be aware of potential bias and misinformation. Outputs from generative AI can be susceptible to bias and misinformation. Always cross-check the information and cite your sources appropriately.
-
Stay curious and informed. As part of our commitment to embracing emerging technologies, you are encouraged to explore these tools, understand how they work, and use them within the parameters set out in this guidance. All members of staff now have access to Bing Chat Enterprise, an AI-powered Copilot for the web available in your Edge Browser. This copilot can help you with various web tasks, such as searching for information or composing emails.
This guidance will be reviewed regularly to address emerging practices and better understanding of the use cases for this technology.
General principles for working with Generative AI
-
Be cautious with sensitive information. Always be mindful of the information you input into generative AI tools. Do not input information that reveals sensitive details or intentions that are not in the public domain.
-
Understand the limitations of generative AI. These tools do not necessarily understand context or bias and can produce different responses to the same question. Their outputs may reflect existing structural inequalities and not consider the need for inclusive practice. Always use your judgement and knowledge to evaluate the outputs.
-
Cite your sources. If you use outputs from generative AI in your work, always cite the tool and any sources used as inputs. Guidance on this is available on the guidance for students page. Also, documents generated by AI as part of your work for St George’s are records and should be treated as such in line with our records management policies and procedures.
Practicalities of using Generative AI
-
Use your university email address. When using generative AI for appropriate uses described in this guidance, you can use your university email address. However, be mindful of the information you input.
-
Explore the technology responsibly. Generative AI is a rapidly evolving field with great potential. However, always remember the ground rules and never input sensitive information or personal data into these tools.
Inappropriate uses of Generative AI
-
Do not use for authoring sensitive messages or summarising sensitive facts. Generative AI should not be used to write papers or reports that reveal sensitive information or the intent of the university that is not yet publicly known.
-
Do not input data for analysis without consent. Generative AI should not be used to analyse data sets without the consent of the data owner.
This interim guidance is intended to help you navigate the use of generative AI in your work. Always remember to use these tools responsibly and ethically, and to respect the privacy and rights of others. If you have any questions or need further clarification, please contact the appropriate department within the university.
FAQs for using generative AI at St George's, University of London
View all Close all
Generative AI refers to artificial intelligence systems capable of generating text, images, or other media in response to prompts. Examples include tools like ChatGPT, ClaudeAI, and Microsoft’s BingChat (also known as Co-Pilot).
Yes, but ensure compliance with relevant St George’s policies and guidance for the use of GenAI (such as records management and data protection. Do not submit any confidential, proprietary, or sensitive information to GenAI tools.
Risks include data privacy breaches, loss of intellectual property rights, and ethical concerns about AI content. Always evaluate these risks before using GenAI tools.
Our current recommendation is for you to use Microsoft’s CoPilot for University’s business, to ensure compliance with privacy and security standards. For use of other tools, please consult the Centre for Technology in Education (CTiE) in the first instance).
No, do not use GenAI tools for authoring sensitive messages or summarising sensitive facts as these tools may retain and reuse data, posing privacy risks.
The ownership of content generated by GenAI is still subject to debate and varies by jurisdiction. In the US, AI-generated content is typically not eligible for copyright protection, as it requires human authorship. In the EU and UK, ownership is possible if there is significant human input, with the UK explicitly extending copyright to computer-generated works where a human made the necessary arrangements. In China, recent court decisions suggest that AI-generated works may be protected if there is human intellectual contribution. Our guidance is for you to always review the specific terms of the AI tool and local laws for detailed guidance.
Be transparent in your use of GenAI by disclosing when/where you have used GenAI tools to generate content, be aware of the biases these tools can harbour, and respect intellectual property. See guidance for disclosure from
Generative AI: guidance for students, also relevant to staff.
Always review, edit, and verify AI-generated content to ensure accuracy and appropriateness. Remember that you are responsible for any AI output you include in your work.
We have published guidance for AI use in assessments for students’ awareness. Please refer
students to the guidance which is available on the website.
Further information
Generative AI in teaching and learning
To find out more about Generative AI in teaching and learning, sign up to the Generative AI in Higher Education MOOC. View access instructions.
Prompting
For information about prompting, view the Generative AI workshop activities:
- Generative AI in Higher Education workshop, block 1
- Generative AI: Implications for assessment workshop, block 3
To access all workshop recordings and resources, please visit our Generative AI Staff programme page.
For more detailed guidelines, contact the Centre for Technology in Education (CTiE).
Relevant policies / resources