Using AI-generated content in assignments
If an assignment brief does not explicitly state that you may use AI-generated content, assume that you cannot paste any AI-generated content into your assignment.
Certain courses may require students to use AI-generated content for certain assignments. In these cases, it will always be explicitly stated in the assignment brief that you may use AI, and the brief will give guidance on how to write a proper declaration and acknowledgment of your use of AI.
Using AI to aid your work
As stated above, generative AIs should be considered as tools supporting research, rather than as places to do research in themselves. For example, you may want to ask for guidance on how to structure an assignment; for feedback on a piece of writing; for clarification or simplification of a complex text; for ideas to “get the ball rolling” on a project or for transcription of spoken word audio or descriptions of images. There are myriad ways AI may be a useful supporting tool, but the important thing to remember is not to use it to “take shortcuts”; if you are using AI to skip reading or to avoid doing work you are finding difficult, then it may not be helping you build the skills and knowledge you need to succeed on your course.
Using AI as a source of information
Please be aware that, as stated above, generative AIs are not considered a legitimate source of information or data. You should not reference or cite them directly unless the brief specifically instructs you to do so. If you use Generative AI to aid you in producing any part of your assessed work, it is your responsibility to cross-reference any information generated by the AI against legitimate sources (e.g. journals, books, online databases, published reports) to make sure that the information you give is up-to-date and accurate. Generative AI often gives convincing, but factually incorrect answers, and should not be depended on for conducting academic research.
It is fundamental to good academic research that you read all your sources yourself, and reference them according to the USW guidelines. USW’s Library and Study Skills teams have a number of resources and workshops to help you understand how to use, evaluate, and reference sources.
Privacy, security, and confidentiality
AIs are not considered a secure platform, and you should not enter any personal or confidential information into an AI system.
|
As with any new technology, you should exercise caution when using generative AI tools including but not limited to Chat GPT, Google Bard and Microsoft Bing. Generative AI can improve the way you work and save time, however, you should be aware of the privacy risks when using such tools. It's worth doing some research before you decide which tool to use, or indeed whether to use them at all, based on how your data is collected and processed. Top tips to safeguard your privacy1. Always check the provider’s privacy policy before use Checking the terms and conditions of the models before using them may seem like a chore but it is well worth the effort. You can find out if the provider uses the data you input to train the model, which could result in any information you input into the chat being used to produce output for other users. In effect, anything you input into the chat may be shared with the provider, who might use it for other purposes. You can find out if the provider is likely to share the information you enter with any third parties and advise on how to exercise your rights under data protection legislation. |
|
2. Keep the data entry to a minimum Avoid entering any personal, private, or sensitive information into a platform where the data may be seen, used or disclosed by the provider. Additionally, we do not know how secure these platforms are - the threat of a data breach or a cyber-attack can never be ruled out. If you do decide to use the tools, only input the minimum amount of data required to achieve your objective. |
|
3. Change the Settings Explore the privacy and security controls inside the AI tools. Most providers make these controls visible and easy to operate within the Settings options. You may be able to delete your chat history and prevent this data being used by the provider.
|
Whilst AIs can be powerful learning aids, they should always be thought of as something to support the research and reading you do by yourself, rather than as somewhere to actually do research. AIs are not considered a legitimate source of information or data for various reasons;
They are usually not able to produce up to date information (for example, ChatGPT is trained only material up to 2021, and cannot give information on anything after that date).
They are not intended to give factually accurate information and often produce convincing, but inaccurate information.
They may not have access to information which is behind a paywall (including a large number of academic articles), and therefore often give quite generalised information (on the level of Wikipedia) rather than the kind of specific, peer-reviewed evidence required for good academic research.
They may reproduce the biases (e.g. sexism, racism) of their training material. Moreover, most AIs were trained on English-language material from US or Western contexts and are extremely limited in terms of their ability to provide reliable information from outside those contexts.
They are programmed to strictly avoid discussion of sensitive issues (e.g. religion, drugs, politics) and may not be able to offer appropriately nuanced or in-depth information on such topics.
AI’s are only trained on existing knowledge, and are less able to produce the kind of novel and original insights which are expected of academic work.