Artificial Intelligence in HDR

Guidelines for the use of Generative Artificial Intelligence in HDR

To maintain academic and research integrity, UNE requires proper acknowledgment of all content produced by generative Artificial Intelligence (AI). It is important to only ever make use of AI tools to inform your research in ethical, critical, and creative ways. Such tools cannot replace your own critical thinking, reasoning, and learning or your original contribution to research. This guide includes important information about the use of generative artificial intelligence that all researchers should be aware of.¹

Generative Artificial Intelligence in Research

"Generative AI can be easily used to create content, including audio, code, text, images and videos, and has transformative potential across a wide range of uses within society and the economy. Its use presents enormous opportunities, as well as many risks and challenges" (Australian Research Council).

Artificial intelligence (AI) uses machine learning to perform functions usually associated with human cognitive functions (e.g. reasoning, learning, problem-solving etc.), responding to a particular set of inputs (McKinsey & Company, 2023; Marr, 2023).

Generative artificial intelligence (GenAI) describes algorithms trained on large data corpora and used to create something new (e.g. text, audio, images, code etc.) in response to a prompt or instruction from a user (McKinsey & Company, 2023; Marr, 2023). Generative AI is becoming widely used in academia and research for:

¹ This content is drawn from HDR601 Candidature at UNE (Module 3: Academic and Research Integrity), which is mandatory research training for all HDR candidates commencing from January 2024.

Should HDR candidates use Generative AI in Research?
In the planning of a research project, HDR candidates are required to consider the ethical, critical, and creative uses of generative AI and to ensure proper citation of all content generated by AI tools and sources that you use.

Failure to reference content from generative AI may be considered academic misconduct or research misconduct, so make sure to avoid the following:

Using output from generative AI tools without acknowledgment,
Presenting false, inaccurate, or misleading information or references generated by AI,
Including substantive content from a generative AI tool into your HDR thesis or publication,
Using GenAI to perform peer review of research or to apply for ethical clearance,
Breaching data privacy, security, copyright, intellectual property, or confidentiality by making sensitive or commercial in confidence material available to Generative AI.

References should provide accurate information for each source and should identify where externally sourced, non-original content has been used in your work. Guides for citing Generative AI are available on The University of Queensland (UQ) webpage.

The use of Generative AI tools in your research project will also depend on:

The nature of your research project
Inputting any research data into AI, is at the risk of losing copyright/intellectual property rights. By putting information into these platforms, you are losing your ability to control the information and how it is used. This could have significant ramifications if you intend to publish data or if you inadvertently release sensitive or highly sensitive data (e.g. personal information, medical information, culturally sensitive information).

Bias and accuracy
In the creation of GenAI tools, there is the potential for bias (embedded during training datasets or in the interpretation of datasets) and inaccurate information (GenAI will not state that it does not know the answer and may generate a false answer, which seems correct). If you choose to utilise GenAI tools during your research, it's important to critically evaluate the information generated and consider how you can verify the authenticity and accuracy of outputs.

Research outputs
If you are considering publishing any of your research, you should look at the information page for authors for each potential journal, to check whether they accept GenAI for article preparation and the terms of GenAI use in publications. The University of Adelaide Library Guide on Artificial Intelligence summarises publisher policies on Using AI in articles for publication.

Originality
The purpose of HDR studies is for you to produce original research. Without knowing the details of the training data and patterns, outputs might be the intellectual property of other researchers (and therefore, not your own work). It is therefore important to reference any GenAI generated information within your publications and/or theses (The University of Queensland has excellent guides to assist with citing Generative AI).

Academic Integrity checks at UNE
Please note: at each milestone review, all documentation and theses are run through the iThenticate software to check for plagiarism and use of Generative AI.

Useful resources

Recommended training
The University of Queensland (UQ) has developed a 30 minute training module on Artificial Intelligence. Please take the time to explore this resource if you are considering the use of GenAI during your candidature.

Recommended reading
Artificial intelligence (2024) Tertiary Education Quality and Standards Agency, Australian Government, https://www.teqsa.gov.au/guides-resources/higher-education-good-practice-hub/artificial-intelligence