Skip to Main Content

Generative AI: For Students

The University of Otago guidelines for use of AI can be found here.

Before using any AI models as part of an assignment CHECK WITH YOUR LECTURER/SUPERVISOR. Each department will have differing guidelines around its use. 

Academic Integrity

Academic integrity means being honest in your studying and assessments. It is the basis for ethical decision-making and behaviour in an academic context. Academic integrity is informed by the values of honesty, trust, responsibility, fairness, respect and courage.

The opposite of this is academic misconduct. Academic misconduct is seeking to gain for yourself, or assisting another person to gain, an academic advantage by deception or other unfair means. The most common form of academic misconduct is plagiarism.

The University of Otago guidelines for Academic Integrity and Misconduct can be found here.

Important considerations

Here are some important considerations that you should keep in mind before using Gen-AI in your studies. 

False Information
As Gen-AI models generally work by predicting language, there is no guarantee that the information provided is correct. If you are using these models to study, always make sure that you evaluate the information provided to thwart misinformation and falsehoods.

Hallucinated Citations
In some cases, Gen-AI models can 'hallucinate' both answers and references. This can be hard to spot as, by nature, Gen-AI models are generally overly confident in their responses. If you are provided with any citation or reference in an Gen-AI model, attempt to find the source elsewhere, the Library Catalogue is generally the best place to start. 

Imbedded bias 
The data sets/algorithms on which Gen-AI models are based, are created by humans, and therefore the biases of the creators are embedded into AI programs. An example of this being in image generator AI showing 'flight attendants' as exclusively female, or 'CEO's as exclusively white and male (source). These biases can reinforce harmful stereotypes, so again, be critical in any information you glean. 

Guidelines for use in Assignments

If your lecturer allows the use of gen-AI in your assignments, here are some guidelines to follow to ensure you are using AI with integrity:

1. Make sure you are not presenting Gen-AI created work as your own.

2. Evaluate any information provided by Gen-AI models for fallacies and bias.

3. Keep up to date with university & industry policies on Gen-AI, keeping in mind they may change.

4. Ensure you are using Gen-AI in an ethical way, both personally and in accordance to academic integrity policies.

5. Referencing is always a part of academic integrity. There is debate around how to do this, however best practice appears to recommend treating the output as a 'personal communication'. 

Who owns the copyright?

Image by Mohamed Hassan from Pixabay

Every Gen-AI model has their own terms of use - be sure to read through them to gain a better understanding on copyright and intellectual property considerations. The University of Adelaide has made a summary of the terms for many major AI tools. If you need clarification, reach out to your Subject Librarian or the University Copyright officer. 

Depending on your use of AI, there are some important things to consider. This is by no means an exhaustive list, and each AI model with have differing practices:

- Any of your research used as a prompt may be used to further train the AI model and it may even become part of its training data set.

- Most AI models are based outside of Aotearoa New Zealand, but your use MUST align with New Zealand laws like Copyright and Privacy. For Copyright: 

  • Section 5(2) has a vague definition of authorship, including from computer generated works, however this can be interpreted in several ways.
  • Also, keep in mind that this act was written prior to the AI boom and is currently in review.

- You can not rely on the Gen-AI outputs to be unique, and it may in fact be plagiarised from already copyrighted works.

When is it safe to use ChatGPT?

Further Reading