Gigamon released its 2026 Hybrid Cloud Security Survey, revealing a fundamental shift in the cyber threat landscape. AI is now involved in 83 percent of reported security breaches, enabling attackers to operate with greater speed and scale than many organizations can defend against.
Despite expanded investments in tools and governance policies, 65 percent of organizations experienced a breach in the past year, reflecting an increase of 40 percent over the past three years. The findings highlight a growing imbalance as adversaries leverage AI to accelerate cyber attacks, while defenders are constrained by fragmented visibility into what’s happening across their networks.
“AI is embedded in nearly every stage of the attack chain, enabling adversaries to outpace detection and response,” said Shane Buckley, President and CEO at Gigamon. “While 93 percent of organizations are investing in new security tools, many still lack visibility into how data moves across their environments, creating confidence without control. Closing this gap requires deep observability, giving security teams the clarity needed to detect threats earlier and respond with precision.”
The annual study, now in its fourth year, surveyed more than 1,000 global Security and IT leaders across Australia, France, Germany, Singapore, the UK, and the US. The survey report, “Reality Check: Exposing the AI Security Illusion,” explores the disconnect between confidence and reality, revealing a critical blind spot in how organizations assess AI-driven risk.
Key Findings: AI is Reshaping Both Offense and Defense
- Confidence is outpacing capability. Nearly two-thirds (64 percent) of organizations believe their ability to secure new AI technologies is “defined” or “integrated,” yet 65 percent experienced a breach in the past year, and 1 in 3 experienced multiple breaches
- AI is transforming both sides of the equation. AI is now embedded across security operations, with 94 percent reporting it autonomously initiates security functions without human interaction, most commonly in alert triage and prioritization (53 percent). At the same time,AI security incidents span multiple risk categories, including:
- ExternalAI attacks (41 percent)
- Internal leaks (30 percent)
- Unsanctioned use of AI (30 percent)
- Direct attacks on LLM systems (33 percent)
- Trust in cloud AI deployments is continuing to erode. As risk increases, data strategies are shifting.Most leaders (72 percent) now believe data lakes are more secure for AI workloads, compared with 70 percent that say they’re reluctant to deploy AI in public cloud environments, up from 54 percent the previous year
- Quantum computing risk is accelerating the timeline. Looking ahead, 87 percent fear “harvest now, decrypt later” attacks, putting today’s encrypted data at future risk and underscoring the longer-term implications of current visibility gaps
Visibility Emerges as the Defining Factor to Combat AI-Driven Threats
Visibility emerged as the top security priority in defending against AI threats in this year’s study, yet it is also where defenders are falling behind. As attackers leverage AI to move faster and operate at scale, organizations still lack a complete view of data in motion across encrypted and East-West traffic, AI workloads, and cloud environments. Among those that experienced a breach, only 30 percent say they had the tools needed to respond effectively, highlighting a critical gap between investment and outcome.
To close this gap, organizations are turning to deep observability. By using network-derived telemetry, including metadata, packets, and flows, and feeding it into security, observability, and cloud tools, organizations can gain complete visibility into all data in motion. Nearly all (93 percent) agreed, reporting that access to packet-level data and rich application metadata is essential to detecting and understanding modern threats, going beyond the visibility that traditional MELT data provides today.
This shift is also reaching the boardroom, with 90 percent of leaders reporting that their boards now support deep observability initiatives, signaling an ongoing commitment to modern approaches that can defend against today’s AI-driven threats.
