The dawn of a new era – AI vs. cybercrime
If you spend enough time reading cybersecurity headlines, you might be forgiven for thinking artificial intelligence (AI) is purely a weapon for the bad guys.
And to be fair, the statistics tell a clear story. Since the rise of generative AI, we’ve seen a staggering 1,200% global surge in phishing attacks.
It’s a topic we’ve covered before at Global Storage, specifically regarding how AI is shaping the future of cybersecurity risks. But focusing solely on AI as a threat vector ignores the other side of the coin. AI could also be the most potent shield we have.
For Australian technology decision-makers, the conversation is shifting from ‘how do we defend against AI?’ to ‘how do we use AI to defend ourselves?’
With 2026 projected to be a pivotal year for autonomous systems and digital sovereignty in our region, leveraging AI for breach response readiness isn’t just a competitive advantage – it’s fast becoming a regulatory necessity.
The autonomous shift in Australia and New Zealand
Change is happening at pace and has been for a while. But technology leaders anticipate that 2026 will bring a transition towards increasingly autonomous AI systems in Australia and New Zealand.
This goes beyond faster chatbots – it’s about creating systems that can reason, plan, and handle security tasks with minimal delay and little need for human intervention.
This shift coincides with stricter regulatory measures driving a stronger convergence between IT and security. In a world where digital sovereignty is a priority, organisations must prove they can detect and neutralise threats instantly, keeping Australian data safe on Australian shores.
Speed is the new compliance currency
Regulatory frameworks in Australia have teeth, and they operate on strict timelines. Consider the Security of Critical Infrastructure (SOCI) Act, which requires reporting significant impact incidents within 12 hours.
Or APRA CPS 234, which demands notification within 72 hours of a material incident.
In the second half of 2024 alone, the OAIC received 595 data breach notifications, with 69% caused by malicious attacks. While 66% of breaches were identified in less than 30 days, that timeline is nowhere near fast enough to meet a 12-hour or 72-hour reporting window.
This is where AI can become your compliance engine. Humans simply cannot sift through terabytes of log data fast enough to identify a patient zero event within 12 hours.
AI, however, excels at this. It enables predictive threat detection and automated response, ensuring that when you do notify the regulator, you have the full picture, not just a guess.
It’s no surprise that 93% of organisations indicate AI will influence their cybersecurity investment decisions over the next year.
Outsmarting the supercharged social engineer
The modern threat actor is no longer sending typo-riddled emails from a ‘prince in Nigeria’. They are using generative AI to create hyper-personalised, error-free campaigns.
Recent reports indicate that AI-powered spear phishing attacks now have a 47% success rate against trained security experts. A notable development is the rise of deepfake business email compromise (BEC). In one instance, a UK engineering firm lost USD $25 million after an employee was duped by a deepfake video conference that mimicked their CFO perfectly.
To embrace proactive cyber defence, we must fight fire with fire. Traditional signature-based detection (looking for known bad code) is useless against a unique, AI-generated email. We need AI-driven behavioural analysis. These tools establish a baseline of normal behaviour for your users – when they log in, what files they access, and how they write emails.
When an account suddenly deviates from that pattern (even if they have the correct password), the AI flags it instantly. It is the difference between finding a breach in 200 days versus 2 minutes.
The necessity of keeping a human in the loop
Despite the power of automation, AI is not a set-and-forget magic wand. It is a force multiplier, not a replacement for human judgment.
Arctic Wolf correctly notes that full automation without oversight is rarely advisable. AI models require fine-tuning to avoid false positives – you don’t want your automated response system quarantining your CEO’s laptop during a board meeting because they logged in from a new iPad.
There is also a trust gap to bridge. Interestingly, research shows that Australians and New Zealanders are ready for AI in critical sectors like emergency response, but only when they are aware of how it is being used. Trust increases significantly with awareness.
The same logic applies to your internal stakeholders. To leverage AI effectively for compliance, you need a strategy that blends algorithmic speed with human strategic oversight.
This ensures your defence is nuanced enough to understand business context, but fast enough to stop a machine-speed attack.
Moving beyond experimental AI
As we dive into 2026, AI in cybersecurity is moving beyond the experimental phase and into full operational maturity.
By integrating AI into your breach response strategy, you aren’t just ticking a box for the SOCI Act or APRA. You are building a resilient organisation capable of withstanding the next generation of threats.
