Skip to Main Content

Generative AI tools for study and research

Generative AI and the Library

A guide on the Library's approach to assessing generative AI tools for study and research

Why do we assess generative AI tools?

The Library reviews AI tools from our publishers and vendors to make sure they’re safe, ethical, and useful for study and research, and that they comply with Murdoch University’s Guiding Principles for Generative AI. These tools are often available in our existing databases and systems, making access to these tools simple and intuitive. 

When can you use generative AI?

All use of generative AI must be within the University's requirements, and it is your responsibility to confirm the level of AI usage allowed. This may differ depending on whether you're a student, researcher, or staff member. Any unauthorised use of generative AI may be regarded as academic misconduct and against university policy. Please see Further support for guidance on allowable AI use.

The Library comprehensively tests gen AI tools for accuracy and minimisation of hallucinations, but these may still occur with any AI tool. Generated content must be critically assessed for accuracy. 

Uploading or sharing any copyrighted materials, such as articles, books, images, etc., or Murdoch intellectual property to AI tools that do not have an existing Murdoch University agreement in place is against the Guiding Principles for Generative AI and in breach of our copyright requirements.

Ethical considerations in generative AI

The Library recognises the complex ethical issues associated with the use of generative AI, particularly in relation to bias, diversity, inclusivity, and environmental sustainability. Our evaluation of generative AI tools carefully considers these concerns to select tools that demonstrate responsible data practices and minimise potential harm.

The publisher-developed tools selected by the Library are governed by content and data management policies designed to uphold fairness and accuracy. These tools draw exclusively on peer-reviewed materials within their own databases and do not incorporate data from third-party or unverified sources. This approach reduces the risk of misinformation and inherent bias.

Environmental sustainability remains a significant area of concern in AI development. The energy and water consumption associated with the use of generative AI is not fully disclosed by leading AI companies, and current assessments rely largely on estimates. There are studies that demonstrate the significant impact generative AI has on power and water usage, and there are also emerging studies that argue AI use is on par with other technologies, such as streaming video and music. The Library encourages informed, considered, and socially responsible decision-making when engaging with generative AI tools.