Scholarship details
Application dates
- Applications close
- 3 May 2026
What you'll receive
You'll receive a stipend of $37,010 per annum for a maximum duration of 1.75 years while undertaking a QUT MPhil. The duration includes an extension of up to three months if approved for your candidature. This is the full-time, tax-exempt rate which will index annually.
You will receive a tuition fee offset/sponsorship, covering the cost of your tuition fees for the first two full-time equivalent years of your masters studies.
As the scholarship recipient, you will have the opportunity to work with a team of leading researchers, to undertake your own innovative research in and across the field.
Eligibility
You need to meet the entry requirements for a QUT Master of Philosophy, including any English language requirements.
Enrol as a full-time, internal student (unless approval for part-time and/or external study is obtained).
You must commence your degree by the start of QUT’s Semester 2 (20th of July 2026).
You must have an honours degree or demonstrated equivalent research expertise in computer science, machine learning, data science, human-computer interaction (HCI) or a closely related field.
You must have relevant experience in high-level programming (e.g. Python) and an understanding of large language model (LLM) architectures and training processes.
A strong desire to work in a socio-technical research environment. Candidates must demonstrate the ability to work with other disciplines and communicate technical concepts to broad audiences (e.g. humanities, arts, and social science colleagues).
How to apply
Apply for this scholarship at the same time you apply for admission to a QUT Master of Philosophy.
The first step is to email Dr Aaron Snoswell
About the scholarship
This MPhil project will investigate the challenge of pluralistic AI alignment – the socio-technical problem of aligning large language model (LLM) behaviour with multiple, potentially conflicting sets of values.
Situated within a two-year Queensland-Bavaria collaborative development grant titled 'Personalised News: Balancing Editorial and Audience Values in AI Alignment', the candidate will work with public service media (PSM) partners to explore how technical architectures can uphold editorial values (such as accuracy and impartiality) while allowing space for individual news audience personalisation.
This project is ideal for a candidate passionate about the future of journalism and the integrity of the public sphere. Tasks may include:
- investigating architectures for pluralistic reward modelling
- evaluating trade-offs between competing alignment methods
- designing and curating real-world datasets in partnership with the ABC (Australia) and BR (Germany)
- training and evaluating new reward models using those datasets.
Key outcomes of the research may include development of a reusable, open-source framework for pluralistic value-aligned news systems and/or publication of findings in premier AI Ethics venues such as ACM Fairness, Accountability and Transparency in AI (FAccT), ACM/AAAI AI, Ethics and Society (AIES), or the Symposium on AI, Ethics, and Society (IASEAI).
Desirable experience for this position includes:
- experience or a strong interest in the news industry, journalism, or public service media
- a proven interest in AI ethics or AI alignment
- published or submitted manuscript/s in AI ethics or computer science conferences
- prior experience running experiments with Linux-based high-performance computing environments or job scheduling systems such as PBS.
The successful candidate will be supervised by Dr Aaron Snoswell and based at QUT in Brisbane, Australia, as part of the GenAI Lab and the Digital Media Research Centre.
This role offers a unique opportunity to work as a technical researcher within a vibrant multi-disciplinary academic environment alongside communication scholars, social scientists, philosophers, and data scientists.