Home Cyber Security

CISOs Reimagine Cloud Security for the AI Age

CISOs Reimagine Cloud Security for the AI Age

Repatriation, resilience, and the rise of metadata-driven defence

When the global shift to cloud computing began more than a decade ago, few could have imagined a day when enterprise leaders might consider moving data back. Yet according to new research from Gigamon, that’s precisely what’s happening as security concerns, data governance, and the growing influence of AI force organisations to rethink how and where they store information.

Gigamon’s report, CISO Insights: Recalibrating Risk in the Age of AI, surveyed over 200 CISOs worldwide, including in Australia. It paints a picture of an industry at an inflection point: one where metadata-driven visibility has become the new cornerstone of cybersecurity, and hybrid cloud strategies are being redefined to meet the demands of an AI-first world.

The Cloud Comes Full Circle

As AI adoption accelerates, so too does the data it generates. AI-related network traffic has nearly doubled in the past year, expanding the potential attack surface and complicating visibility. Three-quarters (75%) of CISOs now believe that public cloud poses greater security risks than any other environment.

In response, many organisations are repatriating data shifting sensitive workloads from public clouds back to private or hybrid infrastructures to regain control. Seventy-three per cent of CISOs say they are actively exploring or implementing this move, citing concerns around compliance, intellectual property, and shadow AI.

This marks a dramatic pivot from the “cloud-first” mantra of recent years. But it’s also a pragmatic one: without full visibility into data in motion, even the most advanced AI-driven defences can be blindsided.

Metadata: The Unsung Hero of Cyber Defence

The survey found 86 per cent of CISOs regard metadata as critical to achieving the level of visibility hybrid environments demand. Metadata allows teams to detect anomalies, reconstruct attack paths, and enhance existing tools all without the heavy storage and cost burden of full packet capture.

As traffic volumes skyrocket, metadata provides the context needed to distinguish signal from noise. It’s no surprise, then, that 52 per cent of CISOs are prioritising metadata integration in 2025–26 to improve security tool effectiveness and cut operational costs.

AI: Both Challenge and Opportunity

AI has emerged as both the greatest accelerator and disruptor of modern cybersecurity. On one hand, adversaries are leveraging AI to automate reconnaissance and generate sophisticated phishing or ransomware campaigns. On the other, CISOs are turning to AI to fill skill gaps, speed up response, and scale their operations.

The research shows 45 per cent of CISOs are already using AI tools to enhance team productivity, while 73 per cent see AI as essential to offset headcount reductions. AI-powered assistants are helping junior analysts make faster, more accurate decisions closing experience gaps and reducing training costs.

Deep Observability: The Next Frontier

The path forward lies in deep observability combining network-derived telemetry, packet intelligence, and log data to create a unified, trusted view of hybrid environments. Eighty-two per cent of CISOs say deep observability is foundational to secure and efficient AI deployment, ensuring that machine learning models operate on high-quality, uncompromised data.

For Australian organisations navigating vast geographies, stringent data sovereignty laws, and tight security budgets, the implications are clear: observability is no longer optional. It’s the new baseline for protecting digital assets in motion.

From Cloud Chaos to Cloud Confidence

The era of “move fast and migrate everything” is giving way to a more measured approach one grounded in visibility, data integrity, and trust. Gigamon’s Deep Observability Pipeline offers exactly that: a unified platform that brings together network-derived telemetry and log data to close blind spots, reduce tool costs, and strengthen security posture across hybrid cloud environments.

As AI continues to redefine what’s possible, one truth remains constant: you can’t protect what you can’t see.