Copyright, authorship and AI

Bond University's position on the use of GenAI tools

The University's position on the uptake of using Generative Artificial Intelligence (GenAI) tools is guided by the need for informed, mindful and critical use of them when creating learning, teaching and research outputs.

Most importantly, it is essential that staff and students refrain from uploading private, sensitive, licenced and copyright-protected materials to any free or paid GenAI services that are known to use these materials for AI training.

The University has endorsed and encourages the use of two GenAI tools for all staff and students that provide licensed access to quality outputs via their Bond login: Microsoft CoPilot and Adobe Firefly.

Copyright considerations

Source materials

Copyright-protected works are often used to rain Generative ArtificiaI Intelligence (GenAI) language models without the permission of copyright owners.

These source materials have been reproduced and stored in open access and commercial genAI platforms. They include text, artworks, photographs, music, audio, computer code and metadata from websites, social media, blogs and scholarly works.

They may also include unauthorised use of files or text you upload or type into an AI platform.

Many GenAI tools do not provide transparency about what sources they are using to train the model, or whether they are storing them, nor which sources are used in generating a specific response to a prompt.

Content generation

Copyright protects original works, not ideas or styles.  

GenAI enables substantial copying of original works which disregards copyright owners' rights and can impact artists' and authors' careers and income.

Australian copyright law

Copyright legislation varies around the world. Users of copyright materials in Australia must abide by Australian Copyright law.

Australia's 'Fair dealing' exceptions are more restrictive than 'fair use' exceptions applied in other countries.

In many countries, including Australia, only humans can be copyright owners. However, that concept is currently being challenged in some jurisdictions.

There is an ongoing international discussion about generative AI, intellectual property and outputs and "whether AI-generated content should be eligible for copyright protection". (WIPO, 2024).

When using GenAI outputs, it is crucial to ensure the information is accurate and the source is genuine. False or plagiarised content can harm the quality and integrity of scholarly work.

Under Australian copyright law, an ‘author’ is:

" a person who creates the work, for example, writing an instruction manual, drawing a graphic, or writing a computer program. For photographs, the author is the person who takes the photograph." 

Source Australian Copyright Council Fact Sheet Ownership of Copyright.

When an original scholarly manuscript is written the copyright automatically belongs to the author, and the author is responsible for the integrity of the work. 

When content generated by an AI tool has been included in a work, the author is still ethically responsible for the integrity of the entire work.

The Committee on Publication Ethics (COPE) has the following position statement on the use of AI in scholarly works:

"Authors who use AI tools in the writing of a manuscript, production of images or graphical elements of the paper, or in the collection and analysis of data, must be transparent in disclosing in the Materials and Methods (or similar section) of the paper how the AI tool was used and which tool was used. Authors are fully responsible for the content of their manuscript, even those parts produced by an AI tool, and are thus liable for any breach of publication ethics."

JAMA (an American health association network platform) also has published a clear statement on the use of AI in the production of scholarly works as well as setting up a JAMA + AI site for health professionals:

"Nonhuman artificial intelligence, language models, machine learning, or similar technologies do not qualify for authorship."

If authors use these models or tools to create content or assist with writing their manuscript, they must take responsibility for the integrity of the content.

STM (The International Association of Scientific, Technical & Medical Publishers) provides concise advice in the white paper Generative AI in Scholarly Communications.

"The paper looks at the ethical, legal, and practical aspects of GenAI, highlighting its potential to transform scholarly communications, and covers a range of topics from intellectual property rights to the challenges of maintaining integrity in the digital age."

Publishers have policies on the use of GenAI in articles submitted for publication, as well as, reviewing and editing processes.

Currently, no major publishers permit AI tools to be an author (Ganjavi, et al., 2024). 

Reasoning includes:

• GenAI is not considered capable of initiating an original piece of research without direction from human authors.
• Plagiarism - there is no guarantee that GenAI content is original and not copied from existing sources. 
• GenAI tools cannot apply the judgement or evaluation added by a researcher or peer reviewer.

The allowable use of GenAI and how it should be disclosed varies between publishers and journals. 

See the list below for some examples of publisher policies.

• Cambridge University Press
• Elsevier
• Emerald
• IEEE
• SAGE
• Science
• Springer Nature
• Taylor & Francis
• Wiley

Some content on this page has been adapted from Charles Sturt University Library. (2024). Using AI tools at university. Charles Sturt University. Retrieved from https://opentext.csu.edu.au/usingai Available under a CC BY 4.0 licence.