Funded PhD Position in Bias in Data Pipelines and AI Systems – University of Queensland, Australia

The University of Queensland (UQ) in Brisbane, Australia, is offering a PhD opportunity focused on bias in data pipelines and artificial-intelligence systems. The project examines how bias can arise, propagate, and be managed across the stages of data collection, preparation, model development, deployment, and user interaction.

This doctoral opportunity is suitable for candidates with backgrounds in computer science, artificial intelligence, data science, machine learning, statistics, information systems, human–computer interaction, or related quantitative disciplines. It is particularly relevant to applicants interested in trustworthy AI, algorithmic fairness, responsible data science, and the societal impact of automated decision systems.

Position Details

• Position: PhD candidate in Bias in Data Pipelines and AI Systems

• University: The University of Queensland

• Research environment: Artificial Intelligence and Machine Learning

• Location: Brisbane, Queensland, Australia

• Research areas: Algorithmic bias, data quality, machine learning, AI fairness, responsible AI, data governance, human–AI interaction

• Contract: Full-time PhD research candidature

• Funding: Funded PhD scholarship / living-stipend support, subject to the official scholarship conditions

• Stipend: Approximately AUD 39,220 per year tax free, according to the aggregated listing

• Application deadline: 7 September 2026, according to the aggregated listing; verify the live UQ project page before applying

Research Focus

Artificial-intelligence systems are often evaluated as though bias originates only in the final predictive model. In practice, bias can enter much earlier through the way data are collected, labelled, sampled, cleaned, integrated, documented, and transformed into training datasets.

The PhD project will investigate how bias moves through data pipelines and AI systems, and how it can be detected, quantified, mitigated, and communicated to users. The research may combine theoretical analysis, empirical machine-learning experiments, data-quality assessment, user studies, and the development of practical tools or frameworks.

Potential research directions may include:

• Identification of bias sources in data collection and labelling.

• Measurement of representation, sampling, and annotation bias.

• Bias propagation through preprocessing and feature engineering.

• Fairness-aware model training and evaluation.

• Robustness of fairness metrics under dataset shift.

• Bias in multimodal or foundation-model pipelines.

• Human interpretation of AI-system errors and uncertainty.

• Documentation, auditing, and governance of datasets and models.

• Trade-offs between predictive performance, fairness, privacy, and transparency.

• Tools for monitoring deployed AI systems over time.

The exact research questions, application domain, datasets, and methodological emphasis should be confirmed in the official University of Queensland project description.

Eligibility and Skills

Applicants should normally hold, or be close to completing, a relevant master’s degree or equivalent qualification. Suitable academic backgrounds may include:

• Computer science or artificial intelligence.

• Data science or information systems.

• Statistics or applied mathematics.

• Machine learning or computational science.

• Human–computer interaction.

• Software engineering or systems engineering.

• Social science or ethics with strong quantitative and computational training.

• A related discipline with demonstrated research experience in AI or data analysis.

Useful preparation may include experience with:

• Python and machine-learning frameworks such as PyTorch or TensorFlow.

• Statistical analysis and experimental design.

• Data management, data cleaning, and reproducible workflows.

• Fairness, explainability, or responsible-AI methods.

• Natural-language processing, computer vision, or multimodal learning.

• Causal inference, uncertainty quantification, or distribution-shift analysis.

• Human-subject research or user studies.

• Data governance and algorithmic auditing.

Applicants should demonstrate strong analytical and communication skills. A competitive application should also show awareness that fairness is context-dependent and cannot be assessed using a single metric without considering the affected population, task, deployment environment, and potential consequences.

Funding and Benefits

The opportunity is listed with a tax-free stipend of approximately AUD 39,220 per year. UQ research projects with confirmed living-stipend scholarships are searchable through the university’s official PhD and MPhil project database.

Applicants should verify whether the advertised stipend includes tuition-fee coverage for domestic and international students, Overseas Student Health Cover where applicable, research support, and any additional scholarship conditions. Australian PhD scholarships often separate living-stipend and tuition components, so the exact offer should be read carefully.

The PhD is expected to be undertaken as a full-time research degree. Candidates should confirm the standard candidature duration, scholarship period, extension rules, commencement date, and eligibility conditions with UQ and the prospective supervisor.

Why Apply?

This PhD is attractive to candidates who want to work on the technical and societal dimensions of modern AI systems. It may be especially suitable for researchers interested in:

• Fair and trustworthy machine learning.

• Responsible deployment of AI in high-impact domains.

• Bias in healthcare, education, employment, or public services.

• Data-centric AI and dataset quality.

• Auditing of foundation models and large-scale AI systems.

• Explainable AI and human-centred evaluation.

• Causal and statistical approaches to fairness.

• Governance and accountability for automated decisions.

The project may also connect with biomedical engineering and medical AI. Researchers working on medical imaging, ultrasound, computational modelling, or clinical prediction could investigate how scanner variation, patient demographics, acquisition protocols, missing data, or annotation practices introduce bias into diagnostic models.

University of Queensland Research Environment

UQ offers research degrees across science, engineering, computing, health, medicine, and social sciences. The university maintains a searchable database of PhD and MPhil projects with confirmed living-stipend scholarships, allowing applicants to filter by programme, research area, and scholarship type.

A UQ doctoral candidate typically works with a supervisory team, completes a research project, undertakes milestone reviews, develops a thesis, and participates in research training and professional development. The precise structure depends on the relevant graduate school, faculty, and doctoral programme.

The project may involve collaboration across computing, information systems, social science, law, ethics, health, or industry. Candidates should examine the supervisor’s publications and research group to understand the intended methodological and application context.

How to Apply

Applicants should first identify the official UQ project and confirm that the advertised scholarship is still available. UQ recommends searching available PhD and MPhil projects with confirmed living-stipend scholarships.

A typical application package may include:

• Academic CV.

• Motivation letter or statement of purpose.

• Master’s degree certificate and academic transcripts.

• Research proposal or response to the advertised project.

• Evidence of programming, statistical, or machine-learning experience.

• Writing sample, thesis, or publication list, if requested.

• Contact details for referees.

• English-language evidence where required.

The application should explain why the applicant is interested in bias in data pipelines and AI systems and identify the skills they bring. Candidates should discuss a specific project, dataset, model, or evaluation problem rather than describing responsible AI only in general terms.

Applicants with biomedical or engineering backgrounds should show how their previous modelling and data-analysis experience can support rigorous fairness evaluation, uncertainty analysis, dataset shift assessment, or clinically meaningful AI validation.

Application Link

Start with the official UQ database of funded research projects:

UQ – Find an available PhD or MPhil project

UQ’s research and doctoral information is also available here:

UQ Research Degrees

For more personalized tips or templates contact our team.

www.applywithmentors.com