Over a relatively short period, artificial intelligence has shifted from standalone experiments and efficiency-focused tools to a technology woven into core business workflows, software platforms, and underlying infrastructure, reshaping how organisations operate and how data travels through their systems. The transfer of that information raises an issue. Data and information form the foundation on which artificial intelligence depends.
Their utility stems from processing extensive data repositories and interfacing with a broad range of applications and external services. The Zero Trust framework was developed to limit and segment access by its foundational principles. The tension between these two forces has emerged as a significant cybersecurity challenge that enterprises are navigating.
Balancing artificial intelligence's need for data access against security requirements for controlled use remains an ongoing challenge. Over the years, cybersecurity teams have focused on shrinking exposure points, isolating network segments, and restricting user access rights. Artificial intelligence, by contrast, produces an entirely different set of conditions.
For AI systems to function effectively, they require access across data stores, application programming interfaces, software platforms, and operational processes. Automated device-to-device communications run continuously in the background, generating layers of operational complexity that conventional security frameworks were never built to handle.
The Zero Trust framework itself has not been identified as the source of the problem. Longstanding assumptions are being brought to the surface by AI. The question has shifted from managing who can retrieve particular data to understanding what those systems actually do with that permission once it is in place.
A number of organisations report confidence in their artificial intelligence security measures. According to Gigamon's 2026 Hybrid Cloud Security Survey, close to 40 percent of respondents reported operating at an integrated level of AI security maturity, with relevant practices embedded throughout their organisations. Violation rates have reached 65 percent, a figure that has risen steadily over the preceding three years, while four in five organisations report that artificial intelligence played a role in recent security breaches.
Funding levels are rising while regulatory structures continue to develop. Yet in a shifting marketplace, those overseeing cybersecurity are not necessarily being directed to expand their technology stack. Attention is directed toward streamlining technology platforms, identifying operational savings, and redirecting resources toward artificial intelligence programmes.
Data breaches across institutions have continued to rise rather than decline. Survey data shows that roughly half of participating organisations recorded a rise in AI-driven insider threat incidents, while over three-quarters identified unauthorised AI usage as a leading obstacle to secure AI deployment. Unauthorised AI adoption within organisations mirrors earlier patterns of unsanctioned technology use, though the potential consequences have been described as more far-reaching.
Traditional network security frameworks were built to handle vertical traffic flows, governing how end users gain access to applications and organisational resources. Machine learning and automation technologies are transforming how industries and institutions operate. AI-driven models, application programming interfaces, and autonomous processes are generating substantial lateral data flows within these computing environments.
Such exchanges are described as fluid, decentralised, and largely undetectable by conventional monitoring systems. Defence teams are often able to identify that an incident occurred, though the underlying cause and method frequently remain unclear. According to Gigamon's study, roughly 74 percent of businesses lack adequate oversight of how AI systems handle data movement across their networks.
Over 40 percent of respondents noted that growing case complexity is extending the time needed to handle incident response. Nearly half of respondents point to expanding blind spots driven by the complexity of cloud environments. Expanding the toolkit further has been described as a progressively harder undertaking.
Drawing greater returns and broader reach from current investments, rather than adding layers of complexity, has been identified as a priority. Taken individually, logs, metrics, and alerts each illuminate different aspects of a system, yet no single source captures the full picture on its own. Analysts in these teams piece together data from isolated platforms, working from inferred patterns rather than real-time visibility into activity.
Analysts suggest that Zero Trust frameworks are likely to remain central to cybersecurity, though practitioners may broaden discussions beyond access control policies alone. Identity confirmation alone is described as insufficient in isolation — analysts indicate organisations are extending authentication processes to encompass how users actually behave. Effective implementation, analysts note, involves visibility into how information flows across the organisation, the ways AI tools interact with that information, and how threat landscapes shift on an ongoing basis.
The focus extends beyond the destination or its practical use, encompassing the flow of data along the way. Data gathered from networks is gaining traction as a valued resource for analysis. According to the study, 92 percent of surveyed organisations consider full visibility into data in motion essential for better security outcomes, while over 90 percent regard deep observability as a foundational requirement for protecting their forthcoming AI deployments.
In an environment shaped by artificial intelligence, observable evidence rather than policy frameworks alone has been identified as central to building trust, with transparency described as an increasingly foundational layer of governance. While Zero Trust remains a broadly adopted security framework, advances in artificial intelligence are altering the nature of the threats it was originally designed to counter. Organisations that prioritise real-time visibility into how data, applications, and AI systems interact across increasingly intricate environments are better positioned to adapt than those relying on fixed controls, analysts have suggested. Analysts have suggested that in an environment shaped by artificial intelligence, openness about how systems operate may prove central to building public trust.
