View Document

Policy on the use of Generative Artificial Intelligence in Research

This is the current version of this document. You can provide feedback on this document to the document author - refer to the Status and Details on the document's navigation bar.

Section 1 - Purpose

(1) The Australian Catholic University (ACU) is committed to the pursuit of knowledge, the dignity of the human person, and the advancement of the common good. In keeping with these values, ACU recognises the transformative potential of generative artificial intelligence (GenAI) in enhancing research excellence and impact. At the same time, the development and use of GenAI in research raise important ethical, social, and legal considerations.

This Policy provides principles-based guidance on the responsible use of GenAI in research conducted by ACU. It ensures that said use complies with ethical standards, University policy, and regulatory requirements, with a view to promoting innovation whilst safeguarding ACU’s research quality and integrity.

Top of Page

Section 2 - Scope / Application

(2) In keeping with the ACU Research Code of Conduct (the “RCoC”), this policy applies to all research activity, and all groups or individuals engaged in research activity (henceforth referred to collectively as “researchers”) conducted under the auspices of, on behalf of, or in collaboration with, ACU. This includes but is not limited to:

  1. academic and professional staff (regardless of employment type);
  2. honorary, adjunct, emeritus and visiting academics;
  3. undergraduate and postgraduate students, including Higher Degree Research students;
  4. staff of other organisations collaborating on ACU research;
  5. partnerships with external organisations, institutions, academics, and other stakeholders
  6. external members of ACU committees;
  7. consultants, independent contractors and external entities; and
  8. volunteers.
Top of Page

Section 3 - Principles

(3) ACU researchers must adhere to the following principles of responsible GenAI use:

  1. Fit for purpose: In performing research tasks and delivering outputs, researchers must give due consideration to the problem they are trying to address, and the various solutions available to them. GenAI tools should only be employed where they are the most appropriate solution, and where they add value to research and support the common good.
  2. Fairness: GenAI solutions are susceptible to error and bias, especially where they are reliant on poor quality data or programming. Accordingly, the use of GenAI in ACU research must include safeguards to manage data quality and data bias risks. This includes, where relevant, ensuring that data inputs and outputs are diverse, inclusive and representative.
    In addition, researchers should seek to critically examine and explicitly address social, cultural and racial biases inherent in GenAI systems, demonstrating how such biases were identified, mitigated, and accounted for in reaching conclusions.
  3. Integrity: The use of GenAI in research must meet ACU’s standards with respect to academic and research integrity, and it must comply with all relevant University policies and regulatory requirements.
  4. Transparency: Researchers must disclose the use of GenAI in research grant applications and disseminated research. As well as building trust, this will facilitate scrutiny and serve as a review mechanism, strengthening ACU research.
    In addition, is it important to acknowledge the unique risks of opaque or complex GenAI systems. Where GenAI processes are not fully explainable or auditable, prudent oversight and transparency becomes even more important.
  5. Accountability: GenAI is a powerful tool with enormous potential, but it is not a substitute for human decision-making. GenAI tasks and outputs must remain subject to researcher oversight, review and intervention, and researchers must be responsible for outputs, regardless of AI use, as required by the Research Authorship Policy.
  6. Environmental Sustainability: Researchers should consider the environmental implications of using GenAI tools, including energy consumption and resource utilisation. Efforts should be made to prioritise sustainable practices, such as employing energy-efficient models, and choosing processes that minimise the carbon footprint associated with computational tasks.
Top of Page

Section 4 - Guidance on Use of GenAI in Research

Literature Reviews

(4) Researchers may use GenAI tools to summarise and compare academic literature but must verify all sources for accuracy and potential bias. GenAI should never replace critical analysis.   Researchers should be aware of, and seek to avoid, any copyright or Intellectual Property violations when using a GenAI tool to summarise literature.

Data Analysis and Interpretation

(5) Researchers must exercise caution when using GenAI for data analysis and only use it where appropriate, Researchers must review outputs and report methodologies transparently. Researchers must be aware of the level of sensitivity and confidentiality of the data they are wishing to analyse before using a GenAI tool.

Thesis and Research Paper Writing

(6) GenAI may be used as a support tool for writing, but it should not replace original thought, critical analysis, or interpretation. Additionally, while the use of GenAI tools for this purpose must be disclosed or acknowledged, GenAI tools should never be listed as an author.

Grant Applications

(7) Researchers must exercise caution when using GenAI for drafting grant applications. The final submission must reflect the researcher's intellectual contributions, and the use of GenAI should be disclosed where required by funding bodies.

Ethics Applications

(8) The use of GenAI to analyse, model or synthesise data collected from research participants gives rise to unique ethical concerns regarding privacy, consent, ownership and reciprocity. Researchers planning to use GenAI for these purposes should give due consideration to the ethical concerns, and seek advice from ACU’s Human Research Ethics Committee, as appropriate.

Indigenous Research

(9) When using GenAI to analyse Indigenous research data, researchers must consult the relevant research ethics and Indigenous research guidance, and uphold Indigenous data sovereignty principles. This consultation ensures that Indigenous data governance processes are followed, traditional knowledge and IP is protected, and community protocols regarding data ownership and technological applications are respected throughout the research process.

Peer Review

(10) You can only use AI in the peer review process where the relevant authority (e.g., publisher, grant agency) does not preclude it, and even then, you should be careful. Staff need to be very confident the GenAI system they use protects confidentiality and Intellectual Property.

Top of Page

Section 5 - Accountability

(11) Violations of this policy may constitute breaches of the RCoC. In addition to adhering to the above-outlined principles and guidance, researchers are expected to:

  • Stay informed about updates in GenAI ethics, best practices, policy, and regulatory requirements.
  • Seek guidance from supervisors, the ACU Research Stewardship Unit, or the ACU eResearch team when unsure about appropriate GenAI use.
Top of Page

Section 6 - Review

(12) An initial ‘light touch’ checkpoint will be conducted six months after the effective date of this Policy.  This interim review aims to assess whether any urgent flaws have emerged due to recent developments or emerging issues in the relevant area. 

(13) A comprehensive review of this Policy must be completed within two years of the effective date, In addition, annual  reviews will be undertaken to ensure ongoing alignment with evolving AI developments and applicable regulatory guidelines.

Top of Page

Section 7 - Further Assistance

(14) Related documents and resources can be found under the Associated Information tab.