New research shows how AI is advancing cybersecurity

Cybersecurity has become a critical concern for anyone who uses the Internet. Cyber threats are becoming increasingly sophisticated, and as such, traditional security measures are often insufficient to protect sensitive data and systems. AI and machine learning are crucial parts of the large cybersecurity systems that defend individuals and organisations from threats.

Torrens University’s Centre for Artificial Intelligence Research and Optimisation (AIRO) is at the forefront of advancing cybersecurity through innovative AI research. With a focus on creating a sustainable and secure digital future, AIRO’s work supports academic excellence while delivering broad societal benefits. By integrating cutting-edge technology with real-world applications, the centre is transforming how cyber threats are detected and mitigated, thereby safeguarding vital systems and personal data.

The latest advancements in AI and cybersecurity research

Threat detection

AI algorithms can analyse vast amounts of data to identify patterns and anomalies that may indicate a cyber threat. Machine learning models are trained on historical data to recognise the signatures of known malware and detect new, previously unseen threats.

In 2024, AIRO researchers published a study into federated learning in intrusion detection systems (IDS) and malware prevention, led by Dr Ammar Alazab. Federated learning enables decentralised AI models to work collaboratively while preserving data privacy, and enhancing real-time threat detection without compromising sensitive information. In this AIRO study, this method was found to improve cybersecurity outcomes and build public trust in emerging digital technologies by adhering to rigorous data protection standards.

Data poisoning

AIRO’s research also addresses the vulnerabilities of AI systems themselves, led by Dr. Kamran Shaukat, a globally recognised top 2% AI cybersecurity expert. By developing robust frameworks to defend against adversarial attacks such as targeted data poisoning, Kamran ensures that AI applications remain reliable and secure.

Data poisoning is when corrupted or misleading data is maliciously introduced into a machine learning model's training dataset, leading to incorrect predictions or classifications, which undermines the security and effectiveness of the AI system affected.

This work has far-reaching implications, bolstering the integrity of AI across sectors that rely heavily on secure data processing and decision-making, including public safety and financial services.

Predictive analytics

AI can predict potential vulnerabilities and future attacks by analysing trends and patterns in cyber threat data. By employing predictive AI, organisations can strengthen their defences before an attack occurs.

AIRO harnesses big data analytics and sophisticated machine learning techniques to uncover hidden patterns of malicious behaviour. This predictive research, led by Professor Tony Jan, plays a critical role in preventing cyber-attacks before they escalate, ensuring that essential services such as financial systems and healthcare networks remain secure and operational. These proactive measures contribute significantly to economic stability and the protection of individual privacy.

The future of AI and cybersecurity research

As AI technology continues to evolve, its role in safeguarding our future will only become more critical, providing a resilient defence against the ever-changing landscape of cyber threats.

Through collaborations with industry leaders such as Palo Alto and Fortinet, government agencies, and other prestigious academic institutions, Torrens University’s AIRO Centre is contributing to a resilient global digital future that ensures the security and well-being of communities worldwide, in alignment with the university’s strategic research directions as well as the Australian government’s initiatives for a secure digital future.

For media enquiries or interviews with our artificial intelligence and cybersecurity experts at AIRO, email media.enquiries@torrens.edu.au.