Agentic hybrid human-AI red teaming orchestration

Australian Catholic University (ACU) is offering a full-time industry-based PhD scholarship to develop next-generation frameworks combining human ingenuity with agentic AI in cybersecurity testing.

This PhD project explores how Agentic AI and human experts can work together in “red teaming”, a process where AI systems are tested for vulnerabilities by simulating adversarial attacks. The research focuses on how to optimally orchestrate collaboration between human creativity and machine autonomy to improve AI safety and trustworthiness. Key areas include: designing intelligent task allocation systems, creating human-in-the-loop interfaces, developing hybrid team performance metrics, and aligning with OWASP LLM Top 10, MITRE ATLAS, and ISO/IEC 42001 standards. The project’s outcomes will help shape global standards for AI assurance, safety testing, and governance, supporting safer GenAI systems in both academic and enterprise settings.

The PhD will be based at ACU’s North Sydney Campus, working closely with ReadyTech’s Emerging Technology Team, which leads applied AI innovation across education, government, and workforce solutions.

How to apply

Applications close: Friday 16 January 2026, 11:59 pm AEDT

Eligibility

• This opportunity is open to both domestic and international applicants.
• The eligibility criteria for the PhD program at ACU can be found in the Higher Degree Research Policy.
• The eligibility criteria for research scholarships can be found in the ACU Research Training Program Scholarship Policy.

Requirements

In addition to the eligibility criteria, the successful applicant should have:

• Strong interest and foundation in AI safety, cybersecurity, and human–AI collaboration; background in computer science/AI/machine learning/information systems/cognitive science.
• Programming proficiency (Python) with at least one ML stack (PyTorch or TensorFlow), and experience in experimental design and evaluation (reproducibility, stats).
• Exposure to agentic systems, e.g., multi-agent/agent frameworks (LangGraph, AutoGen, Semantic Kernel/CrewAI), human-in-the-loop orchestration, or LLM red-teaming methods (prompt attacks, fuzzing).
• Standards and safe practice awareness for GenAI: OWASP Top 10 for LLMs, MITRE ATLAS, NIST AI RMF, ISO/IEC 42001; plus containment/sandboxing (least-privilege tools, isolation, secrets handling) and responsible-AI literacy.

Application process

To be considered for this scholarship, applicants must submit all documentation required for an application to the PhD program at ACU, as detailed in the How to Apply instructions, plus:

• A CV detailing relevant academic and practical experience; and
• A cover letter detailing demonstrated interest and/or experience in the program of research described.

When completing the PhD application form, please indicate that you are also applying for a stipend scholarship and include the name of this project "Agentic hybrid human–AI red teaming orchestration".

Before you apply

Interested candidates should contact Associate Professor Walayat Hussain before applying for the PhD program and scholarship.

walayat.hussain@acu.edu.au