UoW Telstra Grant
Left to right: Dr Thai Vu, Innomente Senior Postdoctoral Researcher; Heath Cooper, CEO Sample Assist; Dr John Le, Lead Investigator
Strengthening AI Security in Occupational Health
Sample Assist and its research partner, Innomente, have joined forces with the University of Wollongong to pioneer a groundbreaking 12-month research collaboration focused on strengthening the security of artificial intelligence (AI) systems used to manage occupational health data.
The project responds to a growing industry need for advanced data protection across safety-sensitive sectors including mining, aviation, construction, transport, maritime, defence, and government, where the integrity of health data directly affects workplace safety and regulatory compliance.
The innovative project explored how Large Language Models (LLMs) can be stress‑tested, broken, and rebuilt to enhance data privacy and security in digital occupational health systems. By identifying and addressing vulnerabilities within AI systems before they are deployed, the research ensures that digital platforms like Sample Assist operate with the highest levels of trust and resilience.
Through this approach, the collaboration demonstrates how responsible AI development can directly improve security for organisations that handle sensitive employee medical information.
The project brought together academic and commercial expertise from Innomente and Sample Assist. Together, the partners have pioneered new methods for securing Large Language Models (LLMs) used in workplace health systems, applying rigorous cybersecurity principles to ensure AI‑enabled health tools meet the demands of regulated, high‑risk sectors.
The project also underscores the importance of applied research partnerships to resolve long standing industry problems and challenges. By aligning cutting‑edge AI research with real‑world operational needs, the project bridges the gap between theory and practice creating solutions that directly improve safety, efficiency, and trust in occupational health management.
To advance the field, the team has published a peer-reviewed article contributing to the wider research effort on securing AI in health technology. The findings provide new frameworks for protecting sensitive data across the sector. In the meantime, Sample Assist is applying these insights to further strengthen its platform architecture and enhance privacy safeguards.