Nearly half of the Australian firms surveyed have replaced entry-level cybersecurity staff with AI systems. According to Barracuda research, 45% of Australian respondents say they cut these roles, compared with 33% who cut senior ones. The figure stands out because globally job cuts hit junior and executive roles almost equally; in Australia, entry-level positions bear the heavier toll.
The research, according to Barracuda, surveyed 2,000 IT and security decision-makers across 16 countries, including 200 Australians. Some 64% of Australian respondents say their professional workforce has shrunk because of AI. These are the company's own figures and should be read as its estimates.
Arctic Wolf offers a concrete example. It claims its AI agents resolve 60% of its customers' security investigations without human intervention. Only 2% of investigations are now handled exclusively by humans, while human-AI collaboration has risen from 1% last year to 38% this year.
The downside is in skills. 68% of respondents admit they lack the capabilities to handle AI-related risks. Barracuda's Siroui Mushegian says AI is changing the landscape faster than teams can build skills, visibility and governance, and argues for a hybrid human-AI model.
The backdrop is rising threats. The Australian Signals Directorate's annual report for 2025-26 counts over 1,500 incident responses, up 20%, and more than 82,000 reports, one every six minutes. Average losses per incident reached $93,500, up 16% year on year. Defence Minister Richard Marles called AI the defining force of the past year's threat landscape.
The consequences show up in data breaches. As early as August, IBM found that 92% of companies affected by AI breaches lacked basic access controls, with average costs of $5.33 million. Office of the Australian Information Commissioner statistics show 667 notifiable breaches in the first half of 2026, nearly equal to the previous period (662). The scale has changed: breaches affecting at least one million people rose to 11, from 4.
The picture is a contradiction. AI is becoming a more useful defensive tool, while its attackers use it to get around access controls. For many firms the problem is not only the technology: without staff able to use it, defence risks being left exposed.
No comments yet. Be the first!