This section illustrates some ethical issues arising from the utilisation of GenAI, and presents the rules adopted by the University of Westminster on the use of GenAI for the preparation of coursework.
Contents of this page:
Can I cause harm by using GenAI? How can I use GenAI ethically?
Consider these areas of concern:
Authorship and plagiarism
GenAI may confuse us as to the boundary between writing generation and editing, but the traditional principles of authorship still apply: 1) You should not pass as yours the work of someone or something else (including GenAI), lest you commit plagiarism. 2) During your studies, you should always submit for assessment work that is your own (or of your group, if the assessment consisted of group work).
Students at the University of Westminster should follow the University guidelines on the use of GenAI (see box on Rules at the University of Westminster Concerning the Use of GenAI by Students).
Bias
As the training data of GenAI is largely taken from the internet, which offers a variety of unhinged opinions, and is created by humans, it will contain inherent human bias, which may be at odds with the values of inclusion, equality and fairness.
Critically evaluate GenAI output bearing this in mind.
Untruth
GenAI output can contain inaccurate or entirely fabricated information (hallucinations) (see tab GenAI: Effective Use).
You should always check what GenAI claims by using other reliable sources not to end up spreading untruths.
Intellectual property
GenAI models are trained on large amounts of human-generated content. Therefore, an individual’s intellectual property may form part of the training data used in these models to subsequently create new content which forms the GenAI tool’s output. However, the original human creator of the content goes without credit or payment for their work, and often no consent for its use is given.
Try to use and acknowledge the original sources.
Also note that it is unclear who owns the copyrights for the outputs of GenAI tools - will it be the person who prompted GenAI or the GenAI company?
Consider this if you expect to make breakthroughs you'd rather be acknowledged for.
Privacy
Public facing GenAI services are able to learn from a user’s prompt in order to further develop their database. Therefore, a user’s prompt is stored by the organisation so any data shared with the tool may later be used in the development of the model, to form part of a future generated output, or be made public in the case of a data breach.
Do not include sensitive information in your prompts.
Note that Grammarly commits not to store users' prompts.
Environment
In order for GenAI tools to process the amounts of data needed for their training, these models need significant quantities of advanced, high-performance chips and extensive server farms that consume large amounts of energy and water.
Be considerate about the amount of interactions you have with AI.
Decline of your cognitive abilities
If you rely extensively on GenAI you may miss an opportunity for deep learning and development of new skills. You may develop dependence on the tool and diminish your cognitive abilities.
Consider the balance of human/AI work that is best for you (see tab GenAI: Effective Use).
Power and control
Our growing reliance on technological devices to perform fundamental cognitive activities makes us dependent on whoever controls these tools (Paglieri, 2024, p. 55).
Consider how to maintain your independence.
More information
For more information on the ethical implications of using GenAI during your studies please have a look at the Generative AI: Ethical Considerations in Higher Education course
The University of Westminster distinguishes between legitimate and illegitimate uses of GenAI by students during their studies. For more information, see the guidance document.
PERMITTED USES of GENAI
At the University of Westminster, AI can be used legitimately as a tool to support learning for example to:
Please note, however, that there may be instances where the uses of GenAI listed above may not be applicable due to specific requirements of some subject areas or individual assignments. Any specific restrictions on using Generative AI will be clearly outlined in the relevant assignment brief, along with instructions provided by your tutor(s).
Very important: any use (from support in planning your work to replication of content) of GenAI included in an assessment submission needs to be disclosed (see our Referencing Guide section on Generative AI and Cite Them Right).
NOT PERMITTED USES of GENAI
At the University of Westminster, the following is NOT permitted:
Inappropriate use of GenAI risks contravening the principles of academic integrity and can be deemed as academic misconduct. Please see table 1 in the Guidance to Students on the Use of Generative Artificial Intelligence for examples of some specific GenAI use cases.
QUESTIONS? GET IN TOUCH!
If you're unsure about using GenAI in your work and have further questions, please contact ai-help@westminster.ac.uk
GUIDANCE DOCUMENTS
Using the GenAI components of Grammarly? Have a look at the guide below that clearly spells out how to use this tool responsibly.
Paglieri, F. (2024) Expropriated Minds: On Some Practical Problems of Generative AI, Beyond Our Cognitive Illusions. Philos. Technol. 37, 55 https://doi.org/10.1007/s13347-024-00743-x