Overview
Many modern AI and machine learning models are “black boxes”: they can make accurate predictions, but we often don’t understand how they arrive at their decisions. This lack of transparency matters because AI is increasingly used in high‑stakes settings such as healthcare, finance, and public policy. Without clear explanations, it can be difficult to trust a model’s output, justify decisions, or identify and correct errors or bias. Many post hoc explainable AI (XAI) techniques have been developed to address this by producing human‑readable interpretations, but these explanations are not always reliable or consistent.
This project explores how we can responsibly use black box models when explanations are limited or uncertain. You will examine how different fields approach this challenge, investigate real‑world decision‑making practices, and consider alternative ways of building trust in AI systems beyond explainability.
Research engagement
You will use a mix of reading, analysis, and discussion to build your understanding. You’ll explore real-world examples and problems through scholarly articles and documents like news articles, reports, and case studies on the topic. The goal is to think critically and consider how we can use AI for decision-making without understanding how the AI works.
Research activities
- Read and summarise case studies of organisations using AI (some successful, some not)
- Identify what helped the AI system work well, or what caused problems
- Use simple ethical frameworks (e.g. fairness, accountability, transparency) to analyse the cases
Research skills
- Critical analysis of real-world technology problems
- Applying ethical thinking to technology
Skills and experience
- Problem-solving and critical thinking skill.
- Reasonable written and verbal communication skills.
- Technical knowledge of AI is helpful but not essential. Bring your willingness to learn.
This project is ideal for students interested in data science and the ethical and societal impacts of AI. No advanced machine learning experience is required, just curiosity and a willingness to engage with big, open-ended questions about how we use technology in the real world.
Start date
2 November, 2026
End date
19 February, 2027
Location
Gardens Point campus (with flexibility for hybrid or online supervision)
Contact
Mythreyi Velmurugan
3138 7838
m.velmurugan@qut.edu.au
