New research from Trend Micro shows the rapid intake of artificial intelligence within cybersecurity capabilities between Australian organizations and growing concerns about new types of cyber risk.
The survey shows that 62% of Australian companies are currently using AI-driven tools within their cybersecurity programs, with an additional 19% considering adopting it. In total, 93% of Australian respondents expressed their willingness to use AI in their cybersecurity objectives.
Adoption and trust
Almost half of respondents (45%) already rely on AI solutions for key processes such as automated asset discovery, risk prioritization, and anomaly detection. Additionally, 38% of Australian organizations surveyed list AI and automation as top priorities for improving cybersecurity practices over the next year.
Speaking about the findings, Andrew Philp, Anz Field Ciso at Trend Micro, stated.
“AI is already transforming how organizations across Australia protect against cyber threats — from faster anomaly detection to automated manuals to time-consuming tasks. But as security teams adopt AI, so does cybercriminals.
The study, conducted by Sapio Research on behalf of Trend Micro, involved 2,250 participants worldwide, with 100 Australian respondents in IT and cybersecurity across a variety of sectors and corporate sizes. This study was completed in March 2025.
New risks on the horizon
Despite their enthusiasm for AI in cybersecurity, a considerable majority of local businesses are predicting new risks. 87% of Australian organizations believe that AI adoption will negatively increase cyber risk exposure within the next three to five years. 43% expect the increase in scale and complexity of AI-driven attacks that require comprehensive changes to their current cybersecurity strategies.
This study highlights several specific areas of risk. These include the potential exposure of sensitive data, the uncertainty about how AI handles and stores unique information, and the risk that reliable models will be misused or avoided. Respondents also pointed to an increasing number of compliance requirements and challenges inherent in monitoring instances of Shadow IT as more endpoints, application programming interfaces (APIs), and the expansion of AI integration.
Live Testing and Vulnerability
The ongoing risk is highlighted by recent results of Trend Micro’s PWN2Own event in Berlin, where the AI category was first introduced. Twelve entries targeted four major AI frameworks and generated snapshots of current AI security vulnerabilities. Nvidia Triton Inference Server was the most frequently targeted platform, but also compromised on Chroma, Redis, and Nvidia Container Toolkits. This can be a single vulnerability exploited to achieve a complete system compromise.
Throughout these frameworks, participants discovered seven unique zero-day vulnerabilities. Affected vendors are given 90 days to address these flaws before further technical details are released.
Seek proactive security
Trend Micro emphasizes the importance of integrating security considerations into AI adoption at every stage. The current environment is characterized by both increased adoption and escalating concerns about risk, presenting challenges for organizations working to balance opportunities and mitigation measures as AI becomes more deeply embedded in the IT environment.
The study concludes that Australian technology and security leaders need to continuously evaluate governance and operational approaches in light of the evolving nature of threats accelerated by AI technology.
