AI in Research
Artificial intelligence (AI), including generative AI, can support many stages of the research lifecycle. It can help researchers work more efficiently, explore data, develop code, analyse text and images, generate synthetic data, and communicate research findings.
Â
AI tools also have important limitations. Their outputs may be inaccurate, incomplete, biased or fabricated, and the use of AI can introduce risks relating to privacy, confidentiality, intellectual property, research integrity and reproducibility.
Researchers remain responsible for how AI is used in their research and for verifying any outputs it produces.
How AI can support research
Depending on the research activity and the tools being used, AI can assist with:
- brainstorming and refining research questions
- searching, summarising and synthesising literature
- writing, reviewing and explaining computer code
- quantitative data analysis and statistical workflows
- qualitative data analysis and coding support
- natural language processing and text analysis
- image, audio and video analysis
- generating synthetic data and supporting data augmentation
- simulations and modelling
- visualisation and interpretation of research data
- drafting, editing and summarising research content
- translating or adapting research communications for different audiences
- automating repetitive research workflows.
AI should generally be used to support researcher judgement rather than replace it.
Working with cutting-edge technology
Take a deep dive into the worlds of AI, Machine Learning and High performance computing.
Browse the resources below to learn more about using artificial intelligence, machine learning, high performance computing and other advanced technologies for your research:
If your research requires other AI tools (not listed below), contact eResearch for assistance:
Before using AI in your research
Before entering research information or data into an AI system, consider:
- Is the tool approved?
Use University-approved applications where required and check whether the application is approved for the type of information you intend to use.
- What information will you provide to the tool?
Do not enter confidential, sensitive, or personally identifiable research information into an AI service unless the service is approved for that information classification and its use is consistent with relevant University requirements. Find the data classification on this Information Security SharePoint Site.
Where appropriate, data should be de-identified before being provided to an AI system, see the eResearch Data De-Identification Guide for details. De-identification does not automatically make information safe to use; consider whether individuals could still be identified when information is combined with other data.
- Does your ethics approval allow it?
Where research involves human participants or their data, ensure your proposed use of AI is consistent with your ethics approval, participant information and consent arrangements.
- Are there contractual or intellectual property restrictions?
Research agreements, data-sharing agreements, confidentiality arrangements, funding conditions or third-party licences may restrict how information can be provided to external AI services (contact Legal Services if unclear).
- Can you verify the output?
AI-generated information should not be assumed to be correct. Researchers should independently verify important claims, calculations, code, references and interpretations.
Microsoft Copilot is approved for the following data classifications. You must log in with your Griffith account to ensure data safety. Data safety is not guaranteed without logging in your Griffith account.
- Unofficial (Public Data)
- Official
- Sensitive
Protected data is NOT approved for use in Microsoft Copilot. Please find the data classification on this Information Security SharePoint Site.
Access URL:Â https://copilot.microsoft.com/
Cost: Free
Other AI Chat Tools: CHATGPT, Claude, Gemini, Grok and other AI chat tools are currently not individually approved, they can still be utilized for research purposes provided they involve Unofficial (i.e., public) data only. This approval is contingent upon adherence to Griffith's criteria for data classification and responsible usage practices. If research necessitates the use of these tools for data classifications beyond Unofficial data, it is crucial to seek guidance from eResearch Services to ensure compliance and proper handling of sensitive information.
Please note: DeepSeek is banned at Griffith University.
AI Productivity Tool:
Microsoft 365 Copilot is an AI-powered productivity tool that integrates with Microsoft 365 apps like Word, Excel, PowerPoint, Outlook, and Teams to enhance creativity, productivity, and skills through real-time intelligent assistance.
Individual Licence Required (Fee Applies)-Â Licence Request Form
AI-Powered Meeting Tool:
Microsoft Teams Premium is an offering that builds on the current Teams experience with additional features that help make Teams meetings even more personalized, intelligent, and protected. It also enables premium capabilities for additional meeting scenarios, AI meeting notes and etc.
Individual Licence Required (Fee Applies) -Â Licence Request Form
 Generative AI is like a creative powerhouse in the tech world. It's all about teaching computers to make new things – art, music, even stories – by learning from existing examples. It's used in designing unique visuals, crafting new music tracks, and even helping scientists discover new drugs. Generative AI is basically a tool for sparking innovation and creativity across different fields by letting machines generate new stuff on their own.
Machine Learning is like giving brains to computers. It's a part of AI that teaches machines to learn from data and make decisions. This tech powers a lot of cool stuff you see daily – from personalized recommendations on streaming platforms to predicting weather or even detecting fraud in banking. It's all about making computers smarter by learning patterns from data.
 Natural Language Processing (NLP) is like teaching computers to understand human language. It's the tech that helps computers read, understand, and even respond to human language just like a person would. Ever used voice assistants like Siri or Alexa? That's NLP in action! It's not just about understanding words; it's also about understanding context and meaning. NLP works behind the scenes in translation apps, chatbots, and even in sorting through tons of text to find specific information. This tech is making it easier for computers to understand us and help us out in a more human-like way.
Computer Vision is like giving eyes to computers. It's all about teaching machines to 'see' and understand the visual world, just like we do. This tech helps computers analyze and make sense of images or videos, which is why it's behind things like facial recognition on your phone or self-driving cars 'seeing' the road. Computer Vision isn't just about recognizing stuff; it's also used in medical imaging, quality control in manufacturing, and even in creating cool Instagram filters! It's making machines see and understand the world around us.
This Fact Sheet provided by Griffith's Legal Services outlines some key legal issues to be aware of when considering the use of Generative AI in relation to research and commercialisation. Read more about managing risks to
- confidential or personal information
- IP rights of third parties
- using generative AI to undertake University work
- implementing generative AI tools
AI is a fast developing technology. Before implementing any AI enabled solutions at the university you should seek advice from the appropriate parts of the University where relevant including:
Getting Advice
Ethical AI is like a set of rules guiding how AI should behave. It's about making sure that Artificial Intelligence is used in a fair, transparent, and responsible way. Imagine AI that respects privacy, doesn't discriminate, and is transparent about how it works. Ethical AI aims to ensure that technology benefits everyone without causing harm or unfairness. It's like setting boundaries and guidelines to make sure AI helps society without creating problems. This means thinking about things like fairness, accountability, and the impact AI has on people's lives. Ethical AI is all about using tech in a way that's good for everyone.
If you'd like more information or are concerned about the ethical use of AI you can contact the teams within Griffith to discuss the topic further regarding your research project.
The Griffith AI Hub (Staff Login Required) is a great resource to learn more about Griffith and AI.
The AI Community of Practice is a place to share and discuss AI news and advancements, to seek collaboration opportunities, to showcase the work Griffith is doing with AI, and to help shape the way we use AI at Griffith.
Please join Griffith's AI Community of Practice to connect, share, and learn with colleagues exploring Artificial Intelligence.
In addition to AI tools that have been specifically approved, AI tools meeting the following criteria are generally approved for use in research involving Unofficial (i.e., public) data only.
Approval Criteria:
| Purpose | Use is limited to research activities only. This approval does not apply to Learning and Teaching or other non-research purposes. |
|---|---|
| Data Sensitivity | Only data classified as Unofficial may be used. Users must review and confirm data classification before using AI tools. See the Information Security site for guidance. |
| Authentication |
|
| Data Management | A documented plan must be in place for data extraction and storage at the end of the project. This plan should align with the required data retention period based on the research impact. |
| Cloud Services | Preference is given to services hosted in Australian data centres. |
| Compliance |
|
| Integrations | Either no system integrations or integrations only with approved cloud-based systems. |
How Griffith researchers are using AI
Griffith researchers are already applying AI across a wide range of disciplines:
- AI technology detects real-time koala crossing in first for field (3 March 2026)
- Australia's first AI in radiology trial secures landmark funding (2 April 2025)
- Training AI to identify ancient artists (17 October 2025)
- Industry Laureate Fellow takes future of farming to next level (11 July 2025)
- Machine learning improves groundwater recharge estimates in WA (19 Dec 2024)
- Spinal Cord Injury rehabilitation technology commercialisation (12 April 2024)
See more examples on the Griffith News website.
Contact us
Speak to our friendly team who are happy to help