The conversation about Artificial Intelligence has been ongoing for months at management committees, but there is a less glamorous reality that is beginning to weigh more than the “demos”: infrastructure. A recent Netskope study portrays an uncomfortable scenario for many organizations: only 38% of Infrastructure and Operations (I&O) leaders believe their infrastructure is fully equipped to handle the new demands brought by AI, while barely 18% express full confidence in having the necessary team and budget to meet future performance, resilience, and security expectations.
This gap, more than a number, describes a growing tension: leadership demands speed, visibility, and results; the teams supporting the technological “backstage” perceive that the environment is not prepared for the role.
Rising expectations… with resources that aren’t growing at the same pace
The report emphasizes a shift in the status of infrastructure: it ceases to be just “maintenance” and becomes a direct business enabler. Four out of five (80%) I&O managers indicate that infrastructure is now central to achieving the organization’s basic goals, and the same 80% state that executive expectations have increased over the past twelve months.
Additionally, the pressure is personalized: 83% perceive that demands on their role have intensified, with 39% describing this increase as “significant.” The problem is that, in parallel, confidence in actual execution capacity does not keep up. In a context where AI raises the bar (capability, latency, data, security, continuity), the study reveals a classic pattern: ambitious goals supported by legacy architectures and fragmented decisions.
AI as a catalyst, but urgent issues have been lingering
Although AI dominates the debate, the picture becomes more nuanced when asked about immediate priorities. I&O leaders place in their “top 3” tasks clearly related to operational continuity and modernization:
- Improve security and performance of remote/hybrid access (43%)
- Increase visibility of network operations and performance (35%)
- Support AI adoption, including agent-based AI (34%)
In other words: AI drives progress, but technical debt (remote access, observability, network) still demands budget and attention. From a tech marketing perspective, this is relevant: many commercial proposals fail when they promise “AI” without addressing the previous bottlenecks that determine whether that AI can be reliably and seamlessly deployed.
The gap with leadership: infrastructure as a “black box”
The report highlights a familiar alignment issue that AI amplifies. Casi two-thirds (63%) of I&O leaders feel disconnected from the strategic conversations shaping IT decisions, and 20% admit not clearly understanding their CEO’s or CIO’s objectives.
This disconnect results in a perception clash: leadership seeks certainty, while technical teams feel pressured to deliver results without early involvement in planning. A particularly revealing statistic: 61% of I&O leaders believe their CEO gets frustrated because infrastructure is not as transparent or understandable as they would like.
Here, AI adds an urgency layer: if infrastructure was already complex (cloud, SaaS, hybrid work), AI introduces new flows (data to models, tool integration, agents, exfiltration controls) and elevates reputational and regulatory sensitivities.
“Performance, resilience, and security”: the triangle where the gap is most evident
The study also captures a direct clash between expectations and perceived reality. Most respondents believe that leadership’s demands regarding:
- performance (55%)
- resilience (58%)
- security (59%)
are unrealistic given current platforms.
Furthermore, although I&O identifies performance and resilience as primary factors leadership expects, the responsible teams do not feel particularly confident in moving the needle securely: only minorities consider themselves “very confident” in managing performance, visibility, cost, security, or resilience (with resilience ranking lowest).
This creates a paradox: proactive modernization is expected, yet changes happen in environments where each adjustment carries risk, and more than half feel that the “as-a-service” model reduces control over infrastructure.
Five moves to “translate” infrastructure into business value
The report offers a pragmatic roadmap to rebuild the link between I&O and the C-suite: it’s not about evangelizing technology, but connecting technical decisions with business outcomes. Among the recommendations, five key areas are highlighted:
- Talk in terms of results: agility, risk reduction, continuity, and productivity instead of purely technological debates (e.g., “ZTNA” as an isolated concept).
- Engage earlier in strategic planning: be involved when decisions about migrations, expansions, new product lines, or AI initiatives are being made, not after commitments are public.
- Advocate for simplicity and consolidation: avoid incremental “patches” that, in the short term, soothe symptoms but increase future complexity and costs.
- Provide continuous visibility: metrics and reporting that make infrastructure understandable for leadership, reducing the perception of a “black box”.
- Position I&O as a facilitator for secure and rapid AI adoption: not a barrier, but ensuring AI does not turn into a source of data leaks, operational risks, or non-compliance.
From a marketing and market perspective, these five areas define where the actual purchase decisions will be made: less about “features” and more about operational credibility (deployment speed, control, impact metrics, and governance capacity in hybrid environments).
Conclusion: AI is no longer just a software debate
The study clearly suggests that the success of AI in business will depend not only on the chosen model but on the organization’s ability to industrialize it without breaking fundamentals. This pushes I&O to adopt a more political and strategic role: translating complexity into decisions, quantifying trade-offs, and building trust around performance, resilience, and security.
For many companies, the immediate challenge is not merely to “adopt AI” as a slogan but to prepare the groundwork so that when a serious use case arrives, infrastructure does not become the bottleneck.
FAQs
What does it really mean for infrastructure to be “AI-ready”?
It means being able to handle new loads (computing, network, data) with guaranteed performance, resilience, and security, as well as having sufficient visibility to govern AI-related tool usage and data flows.
What KPIs help explain infrastructure and AI to the management committee without technical jargon?
Impact metrics: deployment and delivery times, availability and degradations, operational risk (incidents avoided or mitigated), related productivity (time saved), total cost (Opex/Capex), and exposure (sensitive data, access, compliance).
Why does hybrid work remain a priority even with AI pressures?
Because browsers, remote access, and corporate networks still handle most of the actual work. If access is fragile or lacks visibility, implementing AI adds complexity on an unstable foundation.
What is the biggest mistake when modernizing infrastructure for AI?
Investing in isolated initiatives (piecemeal upgrades) without addressing architecture and consolidation: it provides short-term gains but increases complexity, operational costs, and risk just as speed and control are needed most.
Sources:
- Netskope — Crucial Conversations: Line of Sight: Connecting infrastructure decisions to strategic business outcomes