AI Is Replacing Entry-Level Cyber Jobs in Australia, and Skills Gaps Are Growing

Almost half of Australian organisations surveyed say they have replaced or removed entry-level cybersecurity roles with artificial intelligence. The finding comes from Barracuda research reported by Australian technology outlets in late October 2026, and it points to a strange tension in the local security workforce: teams are shrinking at the bottom while struggling to find the skills they need at the top.

The survey covered 2,000 IT and security decision makers at organisations with 100 to 2,000 employees across 16 countries, including 200 in Australia, asking about the previous 12 months. In Australia, 45% said entry-level security roles had been replaced or eliminated with AI, compared with 33% for higher-level roles. Globally, cuts were spread more evenly between junior and senior positions.

The survey also found that 84% of organisations overall said AI was affecting security roles or headcount, and 64% of Australian respondents said their workforce had shrunk because of AI. These are survey results from decision makers, not a census of every employer, but the direction matters for anyone starting out in the field or hiring into it.

A skills gap that runs alongside the cuts

The most striking part of the research is not the job cuts alone. It is what sits next to them. Around 94% of Australian respondents reported cybersecurity skills gaps. The most common gaps were linked to AI-related risks and to advanced identity and phishing threats, each cited by 68% of Australian respondents. Those threats included business email compromise, abuse of OAuth connections, and ways attackers get around multifactor authentication.

Six in 10 said they needed extra expertise in incident response, decision making under pressure, and risk management, and 67% reported needs in AI governance and turning technical risk into business impact. Employers want people who can judge a fast-moving incident, explain risk to a board, and govern AI tools. Those abilities usually grow out of years spent handling smaller incidents first.

Barracuda senior vice president and chief information officer Siroui Mushegian said the findings show how quickly AI is changing the cybersecurity landscape, faster than many security teams can build the skills, visibility and governance to match. She argued that organisations need to reshape security teams by combining AI and automation with human expertise, judgement and visibility, and that the next generation of professionals needs new development pathways and chances for hands-on experience.

Why junior roles are disappearing first

Entry-level security work has always included a large share of routine triage: sorting alerts, checking logs, matching events across systems, and writing up first-pass findings. Those tasks are the easiest for AI tools to absorb, because they follow patterns and draw on large volumes of past examples. Arctic Wolf, cited in coverage of the research, claims its AI agents resolve 60% of its customers’ security investigations without human intervention. That is a vendor claim, not an independent measurement, but it shows how automation is pitched: fewer hands on routine cases, more time for harder problems.

The risk, flagged in Barracuda’s own report, is that removing the bottom rung removes the training ground. Junior analysts learn by working real alerts under supervision. If that work goes to a tool, employers will need another way to build judgement, such as shadowing live incidents and reviewing why an AI recommendation was accepted or overruled.

Pressure inside teams makes the picture harder. Around 52% said security staff were expected to support growing digital operations without extra resources, 44% said business expectations outpaced their teams’ skills and tools, and 24% said their organisations could only protect the business reactively rather than proactively.

Humans still make the final call, for now

One reassuring thread in the research is that human oversight has not vanished. About 65% of organisations said they use a shared decision-making model, in which AI supports investigation and analysis while people stay responsible for context, judgement and final decisions. That balance matters because AI findings still need checking: a tool can link events and suggest a response, but it cannot weigh business impact, legal exposure, or the cost of shutting down a system on a false alarm.

The broader threat backdrop helps explain the squeeze. The Australian Signals Directorate’s 2025-26 annual cyber threat report counted more than 82,000 cybercrime reports, roughly one every six minutes, and more than 1,500 incident responses, up 20% on the previous year. Attackers use AI to move faster and personalise phishing at scale, while defenders use it to spot weak points and respond sooner.

For job seekers, the message is practical rather than bleak. Pure alert triage is shrinking as a career entry point, but demand is growing for people who can validate AI findings, investigate identity-led attacks, respond under pressure, and explain risk in business terms. Courses and certifications still help, but hands-on practice matters more: home labs, capture-the-flag exercises, volunteering for incident simulations, and any role that involves reviewing real alerts with a senior analyst. Employers, in turn, face a choice the report states plainly. Cutting junior roles without building new pathways may save money this year and cost capability later, when there are fewer experienced people to promote into the senior roles that are already hard to fill.

Australia is not alone in this shift, but the local numbers stand out: a 45% entry-level replacement rate, a 94% skills gap rate, and teams that can only react rather than prevent. The organisations that come out ahead will likely treat AI as a tool for junior staff to learn with, not simply a replacement for them.

Frequently Asked Questions

What did the Barracuda survey actually find in Australia?

It found that 45% of Australian respondents said entry-level security roles had been replaced or eliminated with AI, compared with 33% for higher-level roles, and that 94% reported cybersecurity skills gaps. The survey covered 200 Australian decision makers at organisations with 100 to 2,000 employees.

Does this mean entry-level cybersecurity jobs are gone?

No. The findings describe roles replaced or removed at surveyed organisations over the previous 12 months, not the end of junior hiring everywhere. Routine triage work is shrinking as an entry point, while demand is growing for judgement, incident response, identity security and AI governance skills.

What should someone starting in cybersecurity focus on now?

Focus on skills AI cannot easily replace: validating AI findings, investigating identity-based attacks, responding to incidents under pressure, and explaining risk clearly. Hands-on practice through labs, simulations and supervised alert work matters as much as formal study.